P4667Loss of function in PCSK9, atherogenic lipoprotein concentrations, and calcific aortic valve stenosis

Background

Proprotein convertase subtilisin/kexin type 9 (PCSK9) inhibition reduces plasma concentrations of most atherogenic lipoproteins such as low-density lipoprotein cholesterol (LDL-C), apolipoprotein B (apoB) and lipoprotein(a) [Lp(a)]. Atherogenic lipoprotein concentrations have also been linked with calcific aortic valve stenosis (CAVS).

Purpose

1) To determine the association between genetic variants in PCSK9 and lipoprotein-lipid levels, 2) to determine whether loss of function (LOF) in PCSK9 is associated with CAVS and 3) to evaluate if PCSK9 could be implicated in aortic valve interstitial cells (VICs) calcification.

Methods

We built a weighted genetic risk score (wGRS) using 10 single nucleotide polymorphisms at the PCSK9 locus associated with LDL-C in the Global Lipids Genetics Consortium. We determined the association between the wGRS and LDL-C, apoB and Lp(a)] in 9692 participants of the EPIC-Norfolk study using linear regression. We investigated the association between the LOF PCSK9 R46L variant and CAVS risk in a meta-analysis of published (three Copenhagen studies, 1463 cases and 101,620 controls) and unpublished studies (UK Biobank, 1350 cases and 349,043 controls, Malmö Diet and Cancer study, 682 cases and 5963 controls and EPIC-Norfolk, 508 cases and 20,421 controls) prospective, population-based studies using logistic regression adjusted for age and sex. We evaluated PCSK9 expression and localization in explanted aortic valves by capillary Western blot and immunohistochemistry in patients with and without CAVS. Von Kossa staining was used to visualize aortic leaflet calcium deposits. We also assessed VICs calcification potential under oxidative stress condition.

Results

In EPIC-Norfolk, the wGRS was significantly associated with TC, LDL-C, and apoB (all p<0.0001), but not with VLDL-C, HDL-C, triglycerides apoA-I, or Lp(a). Carriers of the R46L variant were at lower CAVS risk (odds ratio=0.71 (95% CI, 0.57–0.88, p<0.001)). Aortic valves of patients with aortic sclerosis (n=12) and CAVS (n=8) presented elevated PCSK9 levels (log2 fold change [FC]=+28.6±5.1, p=0.008 and FC=+39.3±15.2, p=0.02, respectively) compared to controls (n=4).In calcified leaflets, PCSK9 expression co-localized with calcium deposits. PCSK9 expression in VICs was induced by oxidative stress (FC=+2.3±0.4, p=0.02), and subsequent increment in calcification potential was observed.

Conclusion

PCSK9LOF variants are associated with lifelong reductions in non-Lp(a) apoB-containing lipoprotein levels and a lower risk of coronary artery disease and CAVS. PCSK9 is abundant in fibrotic and calcified aortic leaflets. Oxidative stress increases PCSK9 expression in VICs. These results support randomized clinical trials of PCSK9 inhibition in the prevention of CAVS.

Solution structure of carbonmonoxy myoglobin determined from nuclear magnetic resonance distance and chemical shift constraints

Solution NMR structures for sperm whale carbonmonoxy myoglobin have been calculated using 1301 distance restraints determined from nuclear Overhauser enhancement (NOE) measurements on 15N-labeled protein and chemical shift calculations for 385 protons. Starting structures included four crystal forms of myoglobin and 12 structures generated by metric matrix distance geometry. Refinements were also carried out using distance restraints alone. In general, the solution conformations are very close to the crystal structures, although the crystal structures are not consistent with some of the observed NOE connectivities. The solution structures are about as far apart from each other (as measured by backbone root-mean-square deviations) as they are from the crystal conformation. Inclusion of chemical shift restraints both tightened the spread of computed structures (especially in the heme pocket region) and led to structures that were closer to the X-ray conformation. The disposition of the side-chains near the heme group could in many cases be determined with considerable confidence, suggesting that a chemical shift analysis may be a useful adjunct to other sources of structural information available from NMR. In particular, this evidence suggests that the distal histidine residue is slightly displaced from the crystal conformation, but still inside the heme pocket at pH 5.6, that the side-chain of Leu89 is in contact with the heme ring but is probably disordered, and that the heme pocket where ligands bind is virtually identical in solution and in the crystal forms.

Structural determination of the vasoactive intestinal peptide by two-dimensional H-NMR spectroscopy

The structure of the vasoactive intestinal peptide 1-28 in 40% 2,2,2-trifluoroethanol was investigated by two-dimensional 1H-nmr spectroscopy. All 1H resonances, except the gamma, delta, and epsilon protons of the lysine residues, could be sequentially assigned. Numerous intraresidual as well as short-range interresidual nuclear Overhauser effect spectroscopy connectivities were observed. Using a variable-target function minimization, a molecular model consisting of two helical stretches involving residues 7-15 and 19-27 connected by a region of undefined structure was calculated. The existence of an undefined structure between residues 16 and 18 confers mobility to the peptide molecule.

NMR study of the interaction of beta-blockers with sonicated dimyristoylphosphatidylcholine liposomes in the presence of praseodymium cation

The interaction of a series of beta-adrenoreceptor blocking agents with unilamellar dimyristoylphosphatidylcholine (DMPC) liposomes has been studied by proton nuclear magnetic resonance (1H-NMR) in the presence of praseodymium cation (Pr3+) at 30 degrees C. Addition of Pr3+ increased the splitting of the trimethylammonium group signals arising from the phospholipid molecules located at the internal and external surfaces of the bilayers. Adding Pr3+ caused a considerable downfield shift of the external peak but only a slight upfield shift of the internal peak (approximately 3%). The difference in chemical shift of the external and internal peaks (delta Hz) increased linearly as a function of Pr3+ concentration up to 10 mM. The addition of beta-blockers reversed the effect of Pr3+, and propranolol exerted the most pronounced effect, causing complete reversal of the splitting at a concentration of 5 mM. Much higher concentrations of other beta-blockers were required to displace Pr3+. A linear correlation between Pr3+ displacement (P) and logarithm of the apparent partition coefficient (K'm) in DMPC liposomes was obtained for hydrophobic beta-blockers, but hydrophilic beta-blockers did not fit this correlation. It appears that beta-blockers that have ortho or meta substitution require penetration of the liposome bilayers before significant polar group interaction can occur. On the other hand, beta-blockers that have para substitution and low K'm values are able to interact with the polar surfaces of the liposomes without penetration to cause displacement of Pr3+.

Biochemical characterization and osmolytes in papillary collecting ducts from pig and dog kidneys

Papillary collecting duct tubules were prepared in gram quantities from the papillae of dog and pig kidneys. Measurements of substrate and oxygen utilizations by these tubules under both aerobic and anaerobic conditions showed the potential for both glycolysis and oxidative phosphorylation. Oxygen is not necessary to maintain a normal adenosine 5'-triphosphate concentration, but oxidative phosphorylation contributes to more than 65% of the metabolism under aerobic conditions in the two species. Both phosphorus-31 and proton nuclear magnetic resonance spectra recorded from extracts of dog cortex, red medulla, and papilla showed a clear gradient from cortex to papilla for osmolytes, such as glycerophosphorylcholine, sorbitol, inositol, betaine, and sugar phosphates. Other molecules identified in the spectra included glucose, sorbitol, mannitol, lactate, glutamine, alanine, threonine, and adenosine 5'-triphosphate. Conventional biochemical measurements supported these findings. An increase in osmolality from 300 to 600 mosmol/kg H2O for 120 min did not increase the glycerophosphorylcholine and sorbitol concentrations of dog papillary collecting ducts in vitro, but a small effect of a 24-h dehydration was detected in vivo.

The nature of the coenzyme A--manganese(II) complex in solution as studied by proton magnetic resonance

The longitudinal relaxation rates of the adenosine AH(8), AH(2), AH(1′), and the pantetheine PH(3′) protons of coenzyme A (CoA) were measured in solutions of free CoA (0.062 M) and for the CoA-Mn(II) complex. The experiments were carried out in D2O at slightly basic values of the pH at 100.0 MHz and at 298 K. By observing the paramagnetic effects of the weakly bound metal ion on the proton longitudinal rates, distances were calculated between the Mn(II) ion and these four protons. For the complex containing two water molecules, the four distances are 4.8 ± 0.5 Å (1 Å = 0.1 nm) to the AH(8) proton, 6.5 ± 0.6 Å to the AH(2) proton, 6.4 ± 0.6 Å to the AH(1′) proton, and 6.0 ± 0.6 Å to the pantetheine PH(3′) proton. For three water molecules, the four distances are 4.5 ± 0.5, 6.1 ± 0.6, 5.9 ± 0.6, and 5.6 ± 0.6 Å, respectively. The correlation times have been determined previously, based on water proton longitudinal relaxation measurements, and these are 6.6 × 10−10 s for the first case and 4.0 × 10−10 s for the latter case. Experiments were also carried out on 0.010 M CoA solutions and, within experimental error, the distances remained unchanged.These distances, combined with the Mn(II) ion to the phosphorus nuclei distances, together with a knowledge of the conformation of the free CoA in solution are used to construct models for the conformation of the CoA–Mn(II) complex in solution. These results are compared with the available data for the Mn(II)–ATP complex, with the lanthanide ion complexes with dephospho CoA, and with the Co(II) ion complexes with propionyl CoA and the enzyme transcarboxylase.

A magnetic resonance study of the coenzyme A--manganese(II) complex in solution

Proton and 31P magnetic resonance experiments were employed in the study of the weak, stoichiometric, one-to-one complex that is formed between coenzyme A and manganese-(II). The distances in solution between the three phosphorus atoms and the manganese(II) ion are obtained at 287 K and at 298 K. The experiments are carried out over a pH range of 6.3 to 7.7. The correlation time τc of the water molecules bound to the Mn(II) ion in the CoA–Mn(II) complex was calculated using the Solomon–Bloembergen equation from the T1M value obtained for the water protons for solutions where most of the Mn(II) ions are complexed. Paramagnetic spin-lattice relaxation rates for all three phosphorus atoms in the CoA molecule were obtained using proton decoupled Fourier transform 31P nmr techniques. For most probable values of two or three water molecules in the CoA–Mn(II) complex, the distances between the manganese (II) ion and the three phosphorus nuclei are 4.0 ± 0.4 Å (1 Å = 0.1 nm) or 3.7 ± 0.4 Å respectively for 0.062 M CoA solutions. For experiments carried out on 0.010 M CoA, a correlation time which has been determined previously for propionyl CoA was used for the distance calculations and the distances remain unchanged within experimental error. These results indicate that the Mn(II) ion coordinates equally with the oxygens of the three phosphate groups in the CoA–Mn(II) complex.