Nephrol Dial Transplant (2018) 1–8

doi: 10.1093/ndt/gfy287

Downloaded from https://academic.oup.com/ndt/advance-article-abstract/doi/10.1093/ndt/gfy287/5091024 by University of the Western Cape user on 07 September 2018
High-serum phosphate and parathyroid hormone distinctly

ORIGINAL ARTICLE
regulate bone loss and vascular calcification in experimental
chronic kidney disease

Natalia Carrillo-López1,*, Sara Panizo1,*, Cristina Alonso-Montes1, Laura Martı́nez-Arias1, Noelia Avello2,
Patricia Sosa3, Adriana S. Dusso1, Jorge B. Cannata-Andı́a1,4,** and Manuel Naves-Dı́az1,**
1
Bone and Mineral Research Unit, Hospital Universitario Central de Asturias, Instituto de Investigación Sanitaria del Principado de Asturias
(ISPA), REDinREN-ISCIII, Oviedo, Spain, 2Laboratorio de Medicina, Hospital Universitario Central de Asturias, Oviedo, Spain, 3Departamento
de Biologı́a de Sistemas, Universidad de Alcalá, REDinREN-ISCIII, Alcalá de Henares, Spain and 4Departamento de Medicina, Universidad de
Oviedo, Oviedo, Spain

Correspondence and offprint requests to: Jorge B. Cannata-Andı́a; E-mail: cannata@hca.es

*These authors contributed as first authors to this work.
**These authors contributed as senior authors to this work.

ABSTRACT
Background. In chronic kidney disease (CKD), increases in se-
rum phosphate and parathyroid hormone (PTH) aggravate vas-
cular calcification (VC) and bone loss. This study was designed
to discriminate high phosphorus (HP) and PTH contribution
to VC and bone loss.
Methods. Nephrectomized rats fed a HP diet underwent either
sham operation or parathyroidectomy and PTH 1-34 supple-
mentation to normalize serum PTH.
Results. In uraemic rats fed a HP diet, parathyroidectomy with
serum PTH 1-34 supplementation resulted in (i) reduced aortic
calcium (80%) by attenuating osteogenic differentiation (higher
a-actin; reduced Runx2 and BMP2) and increasing the Wnt in-
hibitor Sclerostin, despite a similar degree of hyperphosphatae-
mia, renal damage and serum Klotho; (ii) prevention of bone
loss mostly by attenuating bone resorption and increases in
Wnt inhibitors; and (iii) a 70% decrease in serum calcitriol lev-
els despite significantly reduced serum Fgf23, calcium and renal
24-hydroxylase, which questions that Fgf23 is the main regula-
tor of renal calcitriol production. Significantly, when vascular
smooth muscle cells (VSMCs) were exposed exclusively to high
phosphate and calcium, high PTH enhanced while low PTH at-
tenuated calcium deposition through parathyroid hormone 1
receptor (PTH1R) signalling.
Conclusions. In hyperphosphataemic CKD, a defective sup-
pression of high PTH exacerbates HP-mediated osteogenic
VSMC differentiation and reduces vascular levels of anti-
calcifying sclerostin.
Keywords: gene expression, mineral metabolism, parathyroid-
ectomy, renal osteodystrophy, vascular calcification
INTRODUCTION
The mechanisms underlying the increased risk for vascular cal-
cification (VC) and bone loss in chronic kidney disease (CKD)
remain incompletely characterized. Both secondary hyperpara-
thyroidism (SHPT) and high-serum phosphate (P) are strong
inducers of bone loss and VC, and are associated with poor sur-
vival [1]. Despite the recognized impact of high parathyroid
hormone (PTH) on bone loss, important controversies still exist
regarding a direct role of PTH in VC. While in experimental
CKD high PTH promotes VC independently of P [2], in cul-
tures of bovine vascular smooth muscle cells (VSMCs), PTH 1–
34 concentrations up to 107 M inhibit calcification [3].
Although high-serum P aggravates SHPT, it can also induce
VC directly through PTH-independent mechanisms including
induction of VSMC osteogenic differentiation [4–7]. Because in
CKD high-serum P concurs with elevated PTH, their differen-
tial impact on VC cannot be easily distinguished.
In this study, we discriminated P and PTH contribution to
VC and bone loss using a rat model of CKD with or without
prior parathyroidectomy (PTX), with either normal or high die-
tary P. In vitro studies in rat VSMCs evaluated whether PTH-
driven calcification was direct and the involvement of type I
parathyroid hormone receptor (PTH1R).
MATERIALS AND METHODS
In vivo experimental design
Four-month-old male Wistar rats (n ¼ 38) fed a normal
phosphorus diet (NP) [0.6% P and 0.6% calcium (Ca)] (Panlab,
Spain) were divided into five experimental groups (Figure 1).

C The Author(s) 2018. Published by Oxford University Press on behalf of ERA-EDTA. All rights reserved.
V 1
Microsurgical PTX was performed in two groups of rats with a
sham operation on the remaining rats. Only rats with serum Ca
and intact PTH levels <7.5 and 50 pg/mL, respectively, were
considered parathyroidectomized (92.5% PTX success). Ten
days after PTX, nephrectomy (NX) consisting in the removal of
the right kidney and infarction of 3/4 of the left kidney [8]
was conducted in four groups and a second sham operation was
carried out in the sham group. Two groups of rats were
switched to a high phosphorus diet (HP) (0.9% P and 0.6% Ca).
Simultaneously with the NX, pellets (Innovative Research of
America, FL, USA) providing a continuous physiological infu-
sion rate of 5 mg/kg/day of PTH 1-34 (PTX-PTHsuppl. groups)
or vehicle (Sigma-Aldrich, MO) (NX alone and sham groups)
were subcutaneously implanted. After 14 weeks, rats were placed
in metabolic cages for a 24-h urine collection prior to sacrifice.
All protocols were approved by the Ethics Committee for
laboratory animal of the Oviedo University.
In vitro assessment of VC
R method [3]. The remaining aortic pellet or cells were resus-
The cell line A7r5 (VSMCs from rat aorta, ATCCV, VA) was
grown in dulbecco modified eagle medium (DMEM) supple-
mented with 10% fetal bovine serum (FBS) and 1% penicillin/
streptomycin (Lonza, Belgium) to subconfluence. The growing
medium was replaced by DMEM/F12 (Lonza) with 0.1% bovine
serum albumin (BSA) with either 2 mM Ca and 3 mM P [calcify-
ing medium (CM)] or 1 mM Ca and 1 mM P [non-calcifying
medium (non-CM)] as control. Cells exposed to CM and non-
CM were also exposed to PTH 1-34 (from 1012 M to 106M)
for 4 days, replacing the medium every 48 h in triplicate for each
condition. To test PTH signalling pathways regulating VC down-
stream PTH1R activation, A7r5 were exposed to either a protein
kinase A (PKA) inhibitor (10 mM or 1 mM, H-89, Sigma-Aldrich)
or a protein kinase C (PKC) inhibitor (10 nM or 500 pM,
Staurosporine, Selleckchem, TX) for 4 days under non-CM and
CM, replacing the medium every 48 h, in triplicate per condition.
Cells were collected to obtain total RNA, and to quantify Ca con-
tent and alkaline phosphatase activity (BioAssay Systems, CA).
Analytical and technical procedures
Biochemical markers. Serum and urinary creatinine, Ca
and P and urinary protein were measured using a multi-
FIGURE 1: Experimental design. PTX, parathyroidectomy; SHAM,
sham-operated; PTX-PTHsuppl., parathyroidectomized rats supple-
mented with PTH 1-34.
2 N. Carrillo-López et al.
channel auto analyzer (Hitachi 717). Intact PTH 1-84 and PTH
1-34 were measured using an IRMA or RIA rat PTH assays
(Immutopics, CA and Phoenix Pharmaceutical Inc, CA, respec-
tively). Serum calcitriol was measured by RIA (IDS, UK), and
sKlotho and intact Fgf23 by a sandwich ELISA kit (USCN Life
Downloaded from https://academic.oup.com/ndt/advance-article-abstract/doi/10.1093/ndt/gfy287/5091024 by University of the Western Cape user on 07 September 2018
Science, China and Kainos Laboratories, Japan).
Ex vivo measurements of bone microarchitecture. For tra-
becular and cortical bone microarchitecture, the proximal
metaphysis to mid-diaphysis region of the right tibia were
scanned by micro computed tomography (mCT; SkyScan 1174,
Belgium) and morphometric 2 D and 3 D analysis were per-
formed by the manufacturers provided software (CTAn).
Quantification of Ca content. A frozen aortic fragment
and A7r5 cells were homogenized in 0.6 N NaCl and stirred at
4 C for 24 h.
Upon centrifugation, Ca content was determined colorimet-
rically in the supernatants by the o-cresolphtalein complexone
pended in lysis buffer (125 mM Tris, 2% SDS, pH 6.8) for pro-
tein extraction and quantification. Ca content, normalized for
total protein, was expressed as mg Ca/mg protein.
RNA extraction, cDNA synthesis and quantitative
PCR. Total RNA was extracted using TRI reagent (Sigma-
Aldrich) upon the use of a homogenizer for soft tissues (Omni
International, GA) or fine blending of liquid nitrogen-frozen
fragments of proximal tibias. A 2 mg aliquot of total RNA was
retro-transcribed using the High Capacity cDNA Reverse
Transcription Kit (Applied Biosystems, CA). Quantitative real-
time PCR (qPCR) was performed using pre-developed assays
(Supplementary data, Table S2) with GAPDH as the housekeep-
ing gene. All reactions were performed in triplicate. Relative
gene expression was quantified by the DDCT method [9].
Small interfing RNA of PTH1R. A7r5 cells grown to sub-
confluence were transfected with either siRNA Smart Pool for
the PTH1R gene or with non-targeting siRNA as a negative
control using DharmaFECT transfection reagent (all from
Thermo Scientific). After transfection, cells incubated in CM
were exposed to PTH 1-34 (109 M or 107M) for 4 days.
Statistical analysis
Multivariate generalized linear models (GLMs), t-tests or
Mann–Whitney U tests examined statistical comparisons be-
tween more than two or two experimental groups, respectively.
GLM also examined the independent contribution of high P or
high PTH and their interactions with VC.
Results are expressed as mean6SD or as median and inter-
quartile range. All calculations used the statistical analysis pack-
age SPSS 17.0 (SPPS Inc, IL).
RESULTS
In vivo study
The individual contribution of PTH and P to VC and bone
loss was examined in NX rats fed either high (0.9%) or normal
Table 1. Weight, renal function and biochemical parameters

HP diet NP diet SHAM

(n ¼ 8)
NX PTX-PTHsuppl. þ NX NX PTX-PTHsuppl. þ NX
(n ¼ 6) (n ¼ 8) (n ¼ 8) (n ¼ 8)

Downloaded from https://academic.oup.com/ndt/advance-article-abstract/doi/10.1093/ndt/gfy287/5091024 by University of the Western Cape user on 07 September 2018
Initial weight (g) 377 6 46 386 6 48 414 6 50 400 6 33 419 6 40
Final weight (g) 353 6 56 355 6 68 429 6 48 428 6 30 500 6 58
Creatinine clearance (mL/min) 0.6 (0.2–0.7) 0.5 (0.4–0.8) 1.0 (0.9–1.1)a 0.9 (0.8–1.3)c 2.8 (2.3–3.2)
Serum PTH 1-84 (pg/mL) 15 176 (4064–17 120) Non-detectable 1014 (782–1658)a Non-detectable 761 (372–671)
Serum PTH 1-34 (pg/mL) 84.6 (72.8–120.5) 34.5 (29.6–41.7)a 33.0 (27.2–58.4)a 31.7 (21.9–40.8) 27.0 (25.2–41.2)
Serum calcium (mg/dL) 9.2 6 1.2 6.0 6 0.5a 10.5 6 0.3a 8.1 6 1.3b,c 10.1 6 0.2
Serum phosphorus (mg/dL) 13.6 6 10.8 13.5 6 5.4 4.5 6 0.5a 9.2 6 1.4b 3.9 6 0.4
Serum intact Fgf23 (pg/mL) 1176 (1072–1200) 410 (247–967)a 425 (299–465)a 873 (515–996)b 289 (225–443)
Serum calcitriol (pg/mL) 28.6 6 24.1 8.0 6 3.3 49.9 6 30.9 15.4 6 13.2 44.3 6 22.7
Soluble Klotho (ng/mL) 4.563.0 3.561.5 1.760.8 2.261.6 0.760.2
Phosphaturia (mg/24 h) 67.6 (30.2–76.3) 57.2 (50.8–61.6) 2.6 (0.7–8.1)a 7.1 (2.6–19.9)c 9.3 (8.4–11.6)
Fractional excretion of phosphorus (%) 12.5 6 8.0 7.8 6 5.7 2.0 6 1.8a 2.2 6 1.9c 9.9 6 5.0
Calciuria (mg/24 h) 3.3 6 2.0 2.7 6 1.0 7.6 6 2.0a 3.4 6 1.3b 3.2 6 0.7
Fractional excretion of calcium (%) 0.5 6 0.1 0.7 6 0.5 1.5 6 0.6a 1.1 6 0.9 1.9 6 1.7
Proteinuria (mg/24 h) 98.0 6 34.4 119.8 6 46.4 68.7 6 44.7 68.6 6 95.2 14.5 6 3.3
Values are expressed as mean6SD and median (interquartile range). T-test and Mann–Whitney U test were used as statistical methods. Only rats (n ¼ 38) surviving 14 weeks of the
treatment period were included in the analyses.
a
P < 0.05 versus NX þ HP.
b
P < 0.05 versus NX þ NP.
c
P < 0.05 versus PTX-PTHsuppl. þ NX þ HP.
PTX-PTHsuppl., parathyroidectomized rats supplemented with PTH 1-34; SHAM, sham-operated.

(Figure 2; Table 2). In bone, this HP group showed increased

cortical bone deterioration, with reduced bone volume (BV/
TV), a 30-fold increased cortical porosity (Table 3) and higher
expression of bone formation and resorption genes, more
marked for the latter (50-fold increase), and also with increases
of the Sfrp1 and 4 (Table 4).

Adverse influence of PTH in NX rats fed a HP diet. NX

FIGURE 2: Aortic calcium content in the in vivo experiment. PTX-
PTHsuppl., parathyroidectomized rats supplemented with PTH 1-
34; SHAM, sham-operated. Median (interquartile range) values are
shown. aP < 0.05 versus NX þ HP, bP < 0.05 versus PTX-PTHsuppl.
þ NX þ HP.
(0.6%) P diet. In addition to aortic calcification, skeletal and sys-
temic parameters were compared among all four experimental
groups of NX rats and in sham-operated controls.
Adverse influence of HP diet in NX rats. Uraemic rats fed
a HP diet (NX þ HP group) showed the expected reduction in
renal function and alterations in bone and mineral biochemical
parameters compared with uraemic rats fed an NP diet (NX þ
NP group), including higher serum intact and 1-34 PTH, P,
Fgf23, urinary excretion of Ca and P and lower serum Ca
(Table 1).
In the aorta, the NX þ HP group had higher Ca content
(12.5-fold) and increased expression of bone formation genes
rats fed a HP diet reached the highest serum levels of P, PTH
and Fgf23 (Table 1). The induction of PTX prior to NX in rats
fed HP and supplemented with PTH 1-34 to maintain normal
serum PTH (PTX-PTHsuppl. group) failed to reduce both se-
rum P (NX: 13.5 6 5.4 mg/dL versus PTX-P THsuppl. þ NX:
13.6 6 10.8 mg/dL) and the faster progression of renal dysfunc-
tion (creatinine clearance, 79% lower than sham-operated con-
trols). PTX lowered serum PTH 1-34, Ca and Fgf23 levels
(Table 1) despite a similar high level of serum P.
Furthermore, NX rats fed a HP diet had the highest serum
Fgf23 levels versus all other experimental groups, which
resulted in the highest renal levels of constitutive 24-hydroxy-
lase, the enzyme responsible for calcitriol and calcidiol catabo-
lism (Supplementary data, Figure S1). However, a GLM
analysis demonstrated that PTX and not the HP diet had an in-
dependent contribution to lessen serum calcitriol
(Supplementary data, Table S1).
In the aortas from NX rats fed HP, despite a similarly high-
serum P, PTX reduced Ca content by five times (Figure 2),
through significant decreases in the expression of osteogenic
genes (Runx2 and Bmp2), decreases in vascular contractile a-
actin gene and increases in the Wnt inhibitor sclerostin (Sost)
(Table 2).
In bone from NX rats fed with HP, despite a similarly high-
serum P, PTX attenuated the loss of bone volume and the

Regulation of bone loss and vascular calcification in chronic kidney disease 3

Table 2. Gene expression in aorta

HP diet NP diet SHAM

(n ¼ 8)
NX PTX-PTHsuppl. þ NX NX PTX-PTHsuppl. þ NX
(n ¼ 6) (n ¼ 8) (n ¼ 8) (n ¼ 8)

Downloaded from https://academic.oup.com/ndt/advance-article-abstract/doi/10.1093/ndt/gfy287/5091024 by University of the Western Cape user on 07 September 2018
Runx2 (R.U.) 6.62 (3.14–9.40) 1.71 (1.25–2.69)a 1.73 (0.66–2.01)a 1.07 (0.55–3.44) 0.96 (0.78–1.31)
Bmp2 (R.U.) 5.98 (3.45–7.58) 1.73 (1.36–3.10)a 1.48 (0.62–2.99)a 1.37 (0.56–2.85) 1.10 (0.60–1.20)
a-actin (R.U.) 0.44 (0.20–0.70) 0.84 (0.61–1.75)a 0.37 (0.10–1.30) 0.67 (0.17–1.35) 0.95 (0.64–1.20)
Sfrp1 (R.U.) 1.00 (0.22–1.52) 0.57 (0.54–1.08) 0.20 (0.20–0.47) 0.49 (0.19–0.89) 0.98 (0.93–1.20)
Sfrp4 (R.U.) 0.85 (0.27–2.31) 0.24 (0.16–6.69) 0.22 (0.17–0.87) 0.27 (0.17–0.72) 0.76 (0.69–1.08)
Sost (R.U.) 0.56 (0.55–0.75) 1.79 (0.97–2.65)a 1.12 (0.56–1.17) 0.73 (0.34–1.74)b 1.00 (0.66–1.66)
Values are expressed as median (interquartile range). Mann–Whitney U test was used as statistical method.
a
P < 0.05 versus NXþHP.
b
P < 0.05 versus PTX-PTHsuppl.þNXþHP.
Runx2, Runt-related transcription factor 2; Bmp2, Bone morphogenetic protein 2; Sfrp, Secreted frizzled-related protein; PTX-PTHsuppl., parathyroidectomized rats supplemented
with PTH 1-34; SHAM, Sham-operated; R.U., Relative Units versus SHAM group.

Table 3. Histomorphometric data by mCT in tibia

HP diet NP diet SHAM

(n ¼ 8)
NX PTX-PTHsuppl. þ NX NX PTX-PTHsuppl. þ NX
(n ¼ 6) (n ¼ 8) (n ¼ 8) (n ¼ 8)
Trabecular mCT BV/TV (%) 30.6 6 14.0 23.9 6 1.5 26.8 6 2.5 31.7 6 3.8b 22.8 6 4.1
Tb.Sp (mm) 446 6 269 225 6 28 248 6 67 167 6 21b 248 6 33
Tb.Th (mm) 102.8 6 12.7 83.0 6 4.6a 88.9 6 8.2a 82.1 6 4.7 88.8 6 5.9
Tb.N (mm1) 3.1 6 1.6 2.9 6 0.2 2.9 6 0.3 3.9 6 0.4b 2.6 6 0.4
Cortical mCT BV/TV (%) 56.2 6 11.2 68.5 6 1.8a 71.5 6 1.6a 72.4 6 4.5 72.8 6 2.4
Ct.Th (lm) 660 6 157 804 6 99 785 6 121 931 6 106b 930 6 115
Ct.Po (%) 11.3 6 2.5 0.7 6 0.3a 0.4 6 0.2a 0.4 6 0.2 0.3 6 0.2
Values are expressed as mean 6 SD. T-test was used as statistical method.
a
P < 0.05 versus NX þ HP.
b
P < 0.05 versus NX þ NP.
BV/TV, bone volumen/total volume; Tb.Sp, trabecular separation; Tb.Th, trabecular thickness; Tb.N, number of trabecules; Ct.Th, cortical thickness; Ct.Po, cortical porosity; PTX-
PTHsuppl., parathyroidectomized rats supplemented with PTH 1-34; SHAM, sham-operated.

increases in cortical porosity and also protected the trabecular
bone, as measured by reduced trabecular thickness (Table 3).
These PTH-related bone changes were paralleled by reductions
in the tibial expression of bone formation and resorption genes
as well as in the expression of the Wnt inhibitors Sfrp1 and 4,
with no changes in Dkk1 and Sost genes (Table 4).
Influence of PTH in NX rats fed a NP diet. In uraemic
rats fed with a NP diet, PTX resulted in higher serum P and
higher intact Fgf23 levels, and also in lower serum Ca and cal-
ciuria despite similar serum PTH 1-34 (Table 1).
In NX rats fed NP there were no increases in aortic Ca con-
tent above the levels observed in sham-operated controls
(Figure 2) and PTX also had no effect on aortic Ca.
Accordingly, there were no differences in the aortic expression
of osteogenic, vascular contractile or Wnt pathway inhibitor
genes (Table 2).
In contrast, in the bone of uraemic rats fed NP, PTX im-
proved histomorphometric parameters in both trabecular and
cortical tibia. In fact, bone volume and trabecular number were
higher while trabecular separation was lower (Table 3). Also,
cortical thickness was higher (Table 3). In addition, PTX re-
duced the bone formation gene osteocalcin and the bone re-
sorptive gene cathepsin K (Table 4).
Apart from the important differences observed with high
and low PTH in aortic Ca content and the distinct changes in
systemic and skeletal homeostasis among the four experimental
groups of uraemic rats, a GLM analysis also demonstrated an
independent contribution of a high dietary P to aortic Ca con-
tent (Supplementary data, Table S1). The strong interaction be-
tween hyperphosphataemia and the lack of PTH also supports
that the impact of hyperphosphataemia on aortic Ca content is
significantly different in the presence or absence of PTX
(Supplementary data, Table S1).
In vitro studies
In A7r5 cells exposed to a CM for 4 days, Ca content in-
creased >15 times above the levels in cells incubated in
non-CM (Table 5 and Supplementary data, Figure S2). The
pro-calcifying effect of the CM was attenuated by the addi-
tion of PTH 1-34, at concentrations from 1011 M to 108 M.
The highest protection, which completely eliminated the
stimuli by the CM, occurred with a PTH 1-34 concentration
of 109 M. In contrast, PTH concentrations >107 M, in-
creased A7r5 calcification >60 times above that induced by
the CM (Table 5 and Supplementary data, Figure S2A).
These opposite changes in the calcium deposition induced by
low and high PTH 1-34 were unrelated to differences in cell
survival, as MTT cell viability assays showed >96% viable
cells in all experimental conditions (Supplementary data,
Figure S3).

4 N. Carrillo-López et al.
Table 4. Gene expression in tibia

HP diet NP diet SHAM (n ¼ 8)

NX PTX-PTHsuppl. þ NX NX PTX-PTHsuppl. þ NX

Downloaded from https://academic.oup.com/ndt/advance-article-abstract/doi/10.1093/ndt/gfy287/5091024 by University of the Western Cape user on 07 September 2018
(n ¼ 6) (n ¼ 8) (n ¼ 8) (n ¼ 8)
Runx2 (R.U.) 3.98 (1.76–7.07) 0.30 (0.05–0.77)a 0.47 (0.13–2.25)a 0.12 (0.05–0.51) 0.98 (0.45–1.59)
Osteocalcin (R.U.) 1.21 (0.24–9.71) 0.01 (0.00–0.31) 0.21 (0.08–0.63) 0.01 (0.01–0.07)b 0.74 (0.43–1.57)
Cathepsin K (R.U.) 18.89 (1.52–66.28) 0.12 (0.03–0.92)a 0.38 (0.12–1.56)a 0.07 (0.04–0.22)b 0.95 (0.76–1.30)
Sfrp1 (R.U.) 5.60 (3.09–17.59) 0.49 (0.09–1.59)a 0.87 (0.28–1.67)a 0.34 (0.09–0.86) 1.01 (0.46–1.60)
Sfrp4 (R.U.) 6.86 (2.38–33.12) 0.39 (0.05–1.29)a 0.45 (0.17–1.08)a 0.23 (0.05–0.31) 1.17 (0.48–1.60)
Dkk1 (R.U.) 0.94 (0.75–1.11) 0.76 (0.37–2.09) 0.36 (0.24–2.01) 0.46 (0.06–1.32) 0.99 (0.79–1.56)
Sost (R.U.) 0.55 (0.17–0.69) 0.87 (0.08–1.17) 0.90 (0.27–2.00) 0.62 (0.14–1.32) 1.08 (0.46–2.36)
Values are expressed as median (interquartile range). Mann–Whitney U test was used as the statistical method.
a
P < 0.05 versus NX þ HP.
b
P < 0.05 versus PTX-PTHsuppl. þ NX.
Runx2, Runt-related transcription factor 2; Sfrp, Secreted frizzled-related protein; Dkk1, Dickkopf 1; Sost, Sclerostin; NX, nephrectomized rats; NP, normal phosphorus diet; HP, high
phosphorus diet; PTX-PTHsuppl., parathyroidectomized rats supplemented with PTH 1-34; SHAM, sham-operated; R.U., relative units versus SHAM group.

prevent both the increases in Ca content induced by high PTH

1-34 and the attenuation of Ca deposition driven by low PTH
1-34 (Figure 3).
The high death rate observed in A7r5 cells cultured under
calcifying conditions when exposed to the PKA and PKC inhib-
itors H89 and staurosporin, even at concentrations 10- to 20-
fold lower that the reported optimal for full inhibition [10–11],
have impeded the identification of the pathways involved in the
opposing actions of high and low PTH downstream from
PTH1R activation.

FIGURE 3: PTH signalling for VC trough PTH1R. Upon mock DISCUSSION

transfection (siRNAcontrol) or PTH1R silencing, A7r5 cells were
cultured in non-CM or CM with 109M or 107M PTH 1-34. Data
indicate median (interquartile range) of three independent experi-
ments. aP < 0.05 versus non-CM, bP < 0.05 versus PTH 109M,
c
P < 0.05 versus PTH 107M.
In A7r5 cells exposed to the CM, the increases in total calcium
were paralleled by increases in alkaline phosphatase activity and
Runx2 gene expression, and decreased a-actin gene expression
compared with those in cells exposed to non-CM (Table 5).
According to the dose–response curve, the osteogenic differentia-
tion of A7r5 cells induced by calcifying conditions (2 mM Ca;
3 mM P) was attenuated by the addition of 109 M PTH to the
CM and exacerbated by 107 M PTH. Importantly, in the ab-
sence of calcifying conditions, neither low nor high PTH influ-
enced A7r5 calcification or the levels of osteogenic markers.
To further characterize the pathway for PTH regulation of
VC, we examined the expression of PTH1R and PTH2R in
A7r5 cells. PTH1R mRNA levels were 500-fold higher than
PTH2R mRNA. Furthermore, after a 4-day exposure to calcify-
ing conditions, while PTH1R mRNA levels remained
unchanged, there was an additional 55% reduction of PTH2R
levels. Therefore, the impact of low and high PTH on Ca depo-
sition was measured in A7r5 cells exposed to calcifying condi-
tions upon silencing of the prevalent PTH1R. Even an
incomplete (60%) ablation of PTH1R expression sufficed to
This study, designed to discriminate the relative contribution of
high-serum P and PTH to induce VC and bone loss, has dem-
onstrated a protective role of low PTH and corroborated the ad-
verse impact of high PTH under high dietary P conditions in
in vivo CKD models and also in vitro. Mechanistically, low and
high PTH exert direct and PTH1R-mediated protective and
pro-calcifying actions in VSMC, respectively, which add to the
known contribution of high PTH-driven loss of cortical bone to
VC. More significantly, the use of multivariate analyses and
PTX models of CKD with subcutaneous pellet PTH replace-
ment to normalize serum PTH, has also identified direct effects
of high P and high PTH on VC that are unrelated to the ob-
served abnormalities in the complex and highly controlled Ca/
calcitriol/PTH and P/calcitriol/Fgf23-Klotho axes.
In vivo, the levels of circulating PTH 1-34 achieved using
subcutaneous pellets were similar to those in the sham group.
In PTX rats, the administration of PTH 1-34 failed to normalize
serum Ca [12] possibly due to the independent impact of PTX
on serum calcitriol, unrelated to high serum P (Supplementary
data, Table S1). The use of calcitriol or dietary Ca might have
normalized the hypocalcaemia.
High and low serum PTH distinctly influenced the Ca/calci-
triol/PTH and P/calcitriol/Fgf23-Klotho axes modulating VC
and bone and mineral metabolism in rats fed high or normal di-
etary P. The highest aortic Ca content was observed in the
group of NX rats fed HP concurring with the highest serum
PTH levels, an expected finding from clinical studies

Regulation of bone loss and vascular calcification in chronic kidney disease 5

Table 5. Analysis of PTH effect on VSMC differentiation to osteoblast-like cells
Non-CM
(Control)
Ca content (mg Ca/mg protein) 0.51 6 0.53
Runx2 (R.U.) 1.00 6 0.20
a-actin (R.U.) 1.00 6 0.18
Alkaline phosphatase activity (IU/mL) 3.89 6 0.02
Values are expressed as mean 6 SD of three independent experiments. T-test was used as the statistical method.
a
P < 0.05 versus control (non-CM without PTH) and.
b
P < 0.05 versus CM and.
c
P < 0.05 versus CMþ PTH 109 M.
Ca, calcium; Runx2, Runt-related transcription factor 2; R.U., relative units.
demonstrating an association between high serum PTH and a
higher risk for cardiovascular mortality [13] and also from ex-
perimental studies in uraemic rats with varying concentrations
of PTH [2, 14]. In fact, aortic Ca in NX rats fed high dietary P
was five times higher than in PTX-PTHsuppl. uraemic rats fed
the same high P diet despite similar serum P levels in both
groups, supporting a direct and serum P-independent contribu-
tion of high PTH to worsening VC. Undoubtedly, the develop-
ment of hypocalcaemia upon PTX might attenuate VC. In fact,
although P directly induces VC in vitro even with low Ca, the
degree of VC worsens as Ca concentration rises [15].
Uraemic animals fed a high P diet, with the highest serum
PTH, also presented the most severe decreases in cortical bone
and altered bone microarchitecture. These findings reflect the
catabolic effect of high PTH on bone [16] and the association of
cortical rather than trabecular bone loss with the onset and pro-
gression of VC in uraemic rats [17, 18]. In fact, the aortic gene
expression of markers of bone formation and particularly of
markers of bone resorption was significantly higher in NX rats
fed HP. Both effects were also observed in a similar rat model
of CKD of longer duration (20 weeks) [19]. Nevertheless, the
PTX had an independent contribution to lessening aortic Ca
deposition.
We cannot discard a contribution of low bone resorption to
the reduced VC in the PTX-PTHsuppl. group. Indeed, despite
the limitations of the lCT technique to accurately measure dy-
namic changes in osteoblast and osteoclast number and activity,
in uraemic rats fed a normal P diet, PTX improved both trabecu-
lar bone (higher bone volume and number of trabecules and
lower trabecular separation) and cortical bone thickness.
Furthermore, in uraemic rats fed high P, PTX also attenuated
the marked cortical bone deterioration as reported by other
authors [20].
This study has also corroborated a trend but not a reduction
of tibial Sost by high serum PTH [21]. Indeed, after 14 weeks of
uraemia, tibial Sost mRNA levels were not significantly lower,
in the group of NX rats fed high P and with the highest serum
PTH (Tables 1 and 4). Furthermore, tibial mRNA levels of the
Wnt pathway inhibitors Sfrp1 and Sfrp4 were significantly in-
creased, corroborating that Wnt inhibitors other than Sost con-
tribute to the decreases in bone mineral mass at different CKD
stages, as we and others have reported [19, 21].
Adding to these bone abnormalities associated indirectly
with VC, this study has also identified direct effects of high
PTH on the vasculature: the highest PTH was also associated
6 N. Carrillo-López et al.
Non-CM þ Non-CM þ CM CM þ CM þ
PTH 109 M PTH 107 M PTH 109 M PTH 107 M
0.23 6 0.25 0.18 6 0.21 8.48 6 1.87a 0.43 6 0.12b 33.69 6 17.22a,b,c
14.73 6 5.22a 5.73 6 1.34a,b 51.29 6 18.19a,b,c
Downloaded from https://academic.oup.com/ndt/advance-article-abstract/doi/10.1093/ndt/gfy287/5091024 by University of the Western Cape user on 07 September 2018
1.05 6 0.22 2.00 6 0.71
1.39 6 0.13 1.42 6 0.09 0.42 6 0.08a 1.17 6 0.03b 0.31 6 0.15a,b,c
3.87 6 0.07 3.946 0.03 4.11 6 0.13a 3.93 6 0.03b 4.21 6 0.11a,c
with increased aortic Runx2 and Bmp2 gene expression and
markedly reduced a-actin, a marker of a contractile phenotype
in VSMCs. As reported by other authors [22], the contribution
of P to the osteogenic differentiation of VSMC cannot be dis-
carded. In fact, GLM analysis showed an independent effect of
P on aortic Ca content, with different effects with and without
PTX (Supplementary data, Table S1).
Interestingly, in these uraemic rats fed HP, PTX increased
aorta mRNA levels of Sost, suggesting that the maintenance of
Sost-mediated inactivation of the Wnt pathway may contribute
to the protective effect of low PTH on VC. This finding under-
scores an important clinical aspect of anti-Sost strategies in
CKD, as the use of Wnt activation to preserve bone mass may
aggravate the progression of VC [23–25].
The opposing effects of low and high PTH on VC in uraemia
demonstrated in the in vivo studies were corroborated in vitro.
Dose–response curves for PTH regulation of Ca deposition in
VSMCs demonstrated a protective effect of maintaining PTH
levels below a threshold of 107 M that markedly induced calci-
fication (Supplementary data, Figure S2). Also, a protective
dose of PTH 1-34 (109 M) and a deleterious dose (107 M)
were further characterized, showing opposing effects not only
on Ca deposition, but also on the expression of osteogenic and
vascular contractile genes (Table 5).
Importantly, the adverse impact of PTH 107 M in our study
is in disagreement with the results previously reported by Jono
et al. [3]. These differences may result from the two different
strategies for P-induced calcification (namely b-glycerol phos-
phate by Jono versus 2 mM Ca plus 3 mM P herein), from the
two different types of VSMCs employed (primary bovine cells
and a rat aortic cell line) and also from the use of FBS contain-
ing high levels of calcification inhibitors, such as fetuin-A,
among many others.
In our model, PTH regulated VC mainly through PTH1R.
The silencing of PTH1R, the prevalent PTH receptor in A7r5
cells (whose basal expression remains unchanged upon the in-
duction of osteogenic differentiation by pro-calcifying condi-
tions) partially abolished both the pro-calcifying effect of high
PTH and the attenuation of calcification by low PTH, suggest-
ing that strategies directed to antagonize PTH1R signalling may
help to prevent VC in CKD patients with SHPT. The high mor-
tality rates induced by simultaneous exposure of A7r5 cells to
calcifying conditions and PKA or PKC inhibitors have impeded
the identification of the involvement of either pathway in the
protective or deleterious actions of low and high PTH on VC.
 Our study has also confirmed that, even in the presence of
high serum P, a threshold level for serum Ca is critical to induce
Fgf23 [26]. In fact, although PTH also induces bone Fgf23 syn-
thesis by poorly characterized mechanisms, herein the continu-
ous PTH 1-34 infusion, which raised serum PTH 1-34 to
normal levels, was insufficient to rise serum Fgf23 despite high
serum P. It is possible that in the presence of severe hypocalcae-
mia, high P was insufficient to increase serum Fgf23 levels de-
spite the similar degree of renal failure, in part due to the low
serum calcitriol levels [27], as calcitriol directly induces bone
production of Fgf23 [28] only if there is concurrent hyperphos-
phataemia [29]. It is also possible that the severe hypocalcaemia
caused by reduced intestinal Ca absorption due to low serum
calcitriol directly impaired the positive effect of the vitamin D
hormone on Fgf23 release from bone [30–31].
Importantly, PTX and not high dietary P independently
controlled serum calcitriol despite the increase in the highest
level of 24-hydroxylase, the enzyme responsible for catabolizing
calcitriol in the NX þ HP group (Supplementary data, Figure
S1). This result questions Fgf23 as the main factor responsible
for renal calcitriol synthesis.
Taken together, these findings corroborate the need to ap-
propriately control both serum P and PTH in CKD patients
within optimal ranges to ensure the maintenance of a healthy
bone–vascular axis. Furthermore, it underscores the complex
role of serum levels of PTH, calcitriol and Ca in Fgf23 synthesis
and in the activity of the Wnt/b-catenin pathway in response to
increases in serum P challenging the current paradigm to atten-
uate the progression of renal or vascular damage.
Finally, our results have confirmed for the first time a direct
adverse impact of high PTH, independently of high P on VC in
CKD involving PTH1R signalling.
FUNDING
This work was supported by Plan Nacional de I þ DþI 2008-
2011, and 2013-2016, ISCIII–FEDER (FIS10/0896, FIS13/
00014, FIS16/00637 and FIS17/00715), GRUPIN14-028,
FICYT, IRSIN, FRIAT, RedInRen from ISCIII (RD06/0016/
1013, RD12/0021/1023, RD16/0009/0017, RD12/0021/0006
and RD16/0009/0018). N.C.-L. was supported by ISCIII
(CA10/01327) and GRUPIN14-028, S.P. by Sara Borrell
(CD11/00258) and RenInRen (RD16/0009), C.A.-M. by ISCIII
(PI14/01452), P.S. by FPI-UAH and A.S.D. by Asociación de
Investigación de Fisiologı́a Aplicada.
AUTHORS’ CONTRIBUTIONS
M.N.-D. and J.B.C.-A. designed the study; N.C.-L., S.P. and C.A.-
M. performed experiments; N.C.-L. M.N.-D. and A.S.D. analysed
the data; N.A., L.M.-A. and P.S. made the figures and tables;
N.C.-L., M.N.-D., A.S.D. and J.B.C.-A. drafted and revised the pa-
per; all authors approved the final version of the manuscript.
CONFLICT OF INTEREST STATEMENT
None declared.
Regulation of bone loss and vascular calcification in chronic kidney disease
REFERENCES
1. Block GA, Klassen PS, Lazarus JM et al. Mineral metabolism, mortality, and
morbidity in maintenance hemodialysis. J Am Soc Nephrol 2004; 15:
2208–2218
2. Neves KR, Graciolli FG, dos Reis LM et al. Vascular calcification: contribu-
Downloaded from https://academic.oup.com/ndt/advance-article-abstract/doi/10.1093/ndt/gfy287/5091024 by University of the Western Cape user on 07 September 2018
tion of parathyroid hormone in renal failure. Kidney Int 2007; 71: 1262–1270
3. Jono S, Nishizawa Y, Shioi A et al. Parathyroid hormone-related peptide as
a local regulator of vascular calcification. Its inhibitory action on in vitro cal-
cification by bovine vascular smooth muscle cells. Arterioscler Thromb Vasc
Biol 1997; 17: 1135–1142
4. Cancela AL, Santos RD, Titan SM et al. Phosphorus is associated with coro-
nary artery disease in patients with preserved renal function. PLoS One
2012; 7: e36883
5. Razzaque MS. Phosphate toxicity and vascular mineralization. Contrib
Nephrol 2013; 180: 74–85
6. Lau WL, Linnes M, Chu EY et al. High phosphate feeding promotes mineral
and bone abnormalities in mice with chronic kidney disease. Nephrol Dial
Transplant 2013; 28: 62–69
7. Giachelli CM. Vascular calcification: in vitro evidence for the role of inor-
ganic phosphate. J Am Soc Nephrol 2003; 14: S300–S304
8. Naves-Diaz M, Carrillo-Lopez N, Rodriguez-Rodriguez A et al. Differential
effects of 17beta-estradiol and raloxifene on bone and lipid metabolism in
rats with chronic kidney disease and estrogen insufficiency. Menopause
2010; 17: 766–771
9. Livak KJ, Schmittgen TD. Analysis of relative gene expression data using
real-time quantitative PCR and the 2(-Delta Delta C(T)) Method. Methods
2001; 25: 402–408
10. Chen B, Lin TAO, Yang X et al. Intermittent parathyroid hormone (1–34)
application regulates cAMP-response element binding protein activity to
promote the proliferation and osteogenic differentiation of bone mesenchy-
mal stromal cells, via the cAMP/PKA signaling pathway. Exp Ther Med
2016; 11: 2399–2406
11. Lee SK, Stern PH. Divergent effects of protein kinase C (PKC) inhibitors
staurosporine and bisindolylmaleimide I (GF109203X) on bone resorption.
Biochem Pharmacol 2000; 60: 923–926
12. Neves KR, Graciolli FG, dos Reis LM et al. Adverse effects of hyperphospha-
temia on myocardial hypertrophy, renal function, and bone in rats with re-
nal failure. Kidney Int 2004; 66: 2237–2244
13. Slinin Y, Foley RN, Collins AJ. Calcium, phosphorus, parathyroid hormone,
and cardiovascular disease in hemodialysis patients: the USRDS waves 1, 3,
and 4 study. J Am Soc Nephrol 2005; 16: 1788–1793
14. Ejerblad S, Eriksson I, Johansson H. Uraemic arterial disease. An experi-
mental study with special reference to the effect of parathyroidectomy.
Scand J Urol Nephrol 1979; 13: 161–169
15. Yang H, Curinga G, Giachelli CM. Elevated extracellular calcium levels in-
duce smooth muscle cell matrix mineralization in vitro. Kidney Int 2004; 66:
2293–2299
16. Padagas J, Colloton M, Shalhoub V et al. The receptor activator of nuclear
factor-kappaB ligand inhibitor osteoprotegerin is a bone-protective agent in
a rat model of chronic renal insufficiency and hyperparathyroidism. Calcif
Tissue Int 2006; 78: 35–44
17. De Schutter TM, Neven E, Persy VP et al. Vascular calcification is associated
with cortical bone loss in chronic renal failure rats with and without ovariec-
tomy: the calcification paradox. Am J Nephrol 2011; 34: 356–366
18. Roman-Garcia P, Carrillo-Lopez N, Fernandez-Martin JL et al. High phos-
phorus diet induces vascular calcification, a related decrease in bone mass
and changes in the aortic gene expression. Bone 2010; 46: 121–128
19. Carrillo-Lopez N, Panizo S, Alonso-Montes C et al. Direct inhibition of os-
teoblastic Wnt pathway by fibroblast growth factor 23 contributes to bone
loss in chronic kidney disease. Kidney Int 2016; 90: 77–89
20. Chou FF, Chen JB, Lee CH et al. Parathyroidectomy can improve bone min-
eral density in patients with symptomatic secondary hyperparathyroidism.
Arch Surg 2001; 136: 1064–1068
21. Sabbagh Y, Graciolli FG, O’Brien S et al. Repression of osteocyte Wnt/beta-
catenin signaling is an early event in the progression of renal osteodystro-
phy. J Bone Miner Res 2012; 27: 1757–1772
7
22. Graciolli FG, Neves KR, dos Reis LM et al. Phosphorus overload and PTH
induce aortic expression of Runx2 in experimental uraemia. Nephrol Dial
Transplant 2009; 24: 1416–1421
23. Claes KJ, Viaene L, Heye S et al. Sclerostin: Another vascular calcification
inhibitor?. J Clin Endocrinol Metab 2013; 98: 3221–3228
24. Evenepoel P, D’Haese P, Brandenburg V. Sclerostin and DKK1:
new players in renal bone and vascular disease. Kidney Int 2015; 88:
235–240
25. Moe SM, Chen NX, Newman CL et al. Anti-sclerostin antibody treatment
in a rat model of progressive renal osteodystrophy. J Bone Miner Res 2015;
30: 499–509
26. Quinn SJ, Thomsen AR, Pang JL et al. Interactions between calcium
and phosphorus in the regulation of the production of fibroblast
growth factor 23 in vivo. Am J Physiol Endocrinol Metab 2013; 304:
E310–E320
8 N. Carrillo-López et al.
27. Lopez I, Rodriguez-Ortiz ME, Almaden Y et al. Direct and indirect effects
of parathyroid hormone on circulating levels of fibroblast growth factor 23
in vivo. Kidney Int 2011; 80: 475–482
28. Liu S, Tang W, Zhou J et al. Fibroblast growth factor 23 is a counter-
regulatory phosphaturic hormone for vitamin D. J Am Soc Nephrol 2006;
17: 1305–1315
Downloaded from https://academic.oup.com/ndt/advance-article-abstract/doi/10.1093/ndt/gfy287/5091024 by University of the Western Cape user on 07 September 2018
29. Miedlich SU, Zhu ED, Sabbagh Y et al. The receptor-dependent actions of
1, 25-dihydroxyvitamin D are required for normal growth plate maturation
in NPt2a knockout mice. Endocrinology 2010; 151: 4607–4612
30. Rodriguez-Ortiz ME, Lopez I, Munoz-Castaneda JR et al. Calcium deficiency
reduces circulating levels of FGF23. J Am Soc Nephrol 2012; 23: 1190–1197
31. David V, Dai B, Martin A et al. Calcium regulates FGF-23 expression in
bone. Endocrinology 2013; 154: 4469–4482
Received: 2.4.2018; Editorial decision: 16.7.2018