Deleterious effects of phosphate on vascular and endothelial function via disruption to the nitric oxide pathway

Abstract

Background: Hyperphosphataemia is an independent risk factor for accelerated cardiovascular disease in chronic kidney disease (CKD), although the mechanism for this is poorly understood. We investigated the effects of sustained exposure to a high-phosphate environment on endothelial function in cellular and preclinical models, as well as in human subjects.

Methods: Resistance vessels from rats and humans (± CKD) were incubated in a normal (1.18 mM) or high (2.5 mM) phosphate concentration solution and cells were cultured in normal- (0.5 mM) or high-phosphate (3 mM) concentration media. A single-blind crossover study was performed in healthy volunteers, receiving phosphate supplements or a phosphate binder (lanthanum), and endothelial function measured was by flow-mediated dilatation.

Results: Endothelium-dependent vasodilatation was impaired when resistance vessels were exposed to high phosphate; this could be reversed in the presence of a phosphodiesterase-5-inhibitor. Vessels from patients with CKD relaxed normally when incubated in normal-phosphate conditions, suggesting that the detrimental effects of phosphate may be reversible. Exposure to high-phosphate disrupted the whole nitric oxide pathway with reduced nitric oxide and cyclic guanosine monophosphate production and total and phospho endothelial nitric oxide synthase expression. In humans, endothelial function was reduced by chronic phosphate loading independent of serum phosphate, but was associated with higher urinary phosphate excretion and serum fibroblast growth factor 23.

Conclusions: These directly detrimental effects of phosphate, independent of other factors in the uraemic environment, may explain the increased cardiovascular risk associated with phosphate in CKD.

Keywords: cardiovascular risk; chronic kidney disease; endothelial function; nitric oxide; phosphate.

© The Author 2016. Published by Oxford University Press on behalf of ERA-EDTA. All rights reserved.

Figures

FIGURE 1

Rat mesenteric vessels incubated in a high-phosphate concentration (2.5 mM) solution have impaired endothelium-dependent and -independent vasodilatation, an effect that can be reversed when the vessels are co-incubated with a phosphodiesterase-5 inhibitor (PDE5I), zaprinast. Vessels were incubated for 16 h in either a normal (1.18 mM) or high (2.5 mM) phosphate concentration solution and all responses are expressed as mean ± SEM. (A and C) Vasodilatation to increasing concentrations of carbachol or SNP, respectively, expressed as a % of maximal contraction with PEP 1 × 10−5 M. (B) Contraction with PEP in the presence and absence of L-NAME. (D and E) Vasodilatation to increasing concentrations of carbachol or SNP, respectively, in the presence of zaprinast (PDE5I) expressed as a % of maximal contraction with PEP 1 × 10−5 M. n = 10–13 vessels (see Table 1).

FIGURE 2

Human resistance vessels from patients with (C) and without (A) CKD incubated in a high-phosphate concentration (2.5 mM) solution have impaired endothelium-dependent vasodilatation. Endothelium-independent vasodilatation is impaired in vessels from humans without CKD (B) but preserved in those with CKD (D). Vessels were incubated for 16 h in either a normal (1.18 mM) or high (2.5 mM) phosphate concentration solution and all responses are expressed as mean ± SEM. (A and C) Vasodilatation to increasing concentrations of carbachol expressed as a % of contraction with PEP 1 × 10−5 M in vessels from living kidney donors without CKD (A) and from patients with CKD stage 5 (C). (B and D) Vasodilatation to increasing concentrations of SNP expressed as a % of contraction with PEP 1 × 10−5 M in vessels from living kidney donors without CKD (B) and from patients with CKD stage 5 (D). n = 9–14 vessels (see Table 3).

FIGURE 3

HUVECs express reduced total (A) and phospho (B) eNOS and increased nitrotyrosine (C) when cultured in high-phosphate concentration medium (3 mM). PKG expression is reduced in rat VSMCs cultured in high- compared with normal-phosphate concentration medium. **P < 0.05. HUVECs and VSMCs were cultured in 0.5 mM or 3 mM phosphate concentration medium. Ten µg of protein was fractionated on SDS page gels. Primary antibodies for Total eNOS (1:1000), Phospho eNOS (1:200) (both Cell Signalling Technology), nitrotyrosine (1:500, R & D Systems), PKG (1:200, Enzo Life Sciences) and protein expression normalized to GAPDH. Densitometry was performed. In (A)–(D), top panels show representative gel from cells cultured in 0.5 or 3 mM phosphate concentration medium and bottom panels show GAPDH. The panels were all taken from a single gel and each band represents one well from separate six-well plates.

FIGURE 4

cGMP concentration is significantly reduced in rat vessels incubated for 24 h in a high-phosphate concentration solution (A), whereas the intracellular calcium level in HUVECs is unchanged in a high-phosphate environment (B). (A) Rat mesenteric resistance vessels were incubated in a 1.18 mM (n = 11) or 2.5 mM (n = 11) phosphate concentration solution for 24 h. Results are expressed as cGMP pM/µg protein. Analysis was by the Mann–Whitney U test. (B) Cells were plated in glass-bottomed plates and cytosolic loading of FURA-2-AM was achieved by incubating the cells with FURA-2-AM. Values are mean ± SEM and are from at least four experiments with the number of cells included indicated by (n).

FIGURE 5

Change in FMD from baseline visit measures, following intervention with lanthanum and phosphate expressed as a % change from the baseline visit.

FIGURE 6

Possible mechanism of action of phosphate as a cardiovascular risk factor causing both a pro-inflammatory state and endothelial dysfunction. Using evidence presented in this manuscript and previous studies, we propose this as the possible mechanism of action of phosphate resulting in endothelial dysfunction and a pro-inflammatory state. Additionally, we have some evidence of an effect of phosphate on cell growth (data not yet published).