The detailed list of materials (company name, catalog number, sequences, etc.) can be found in the “Resources table” in the Supplementary Material.

HAoSMCs were maintained in Dulbecco’s Modified Eagle Medium (DMEM) supplemented with 10% fetal bovine serum (FBS), antibiotic antimycotic solution, sodium pyruvate, and L-glutamine. Cells were maintained at 37°C in a humidified atmosphere with 5% CO2. Cells were grown to ∼90% confluency and used between passages five and 8. To induce calcification, HAoSMCs were exposed to an osteogenic medium (OM) that was obtained by supplementing the growth medium with inorganic phosphate (Pi) (NaH₂PO₄-Na₂HPO₄, 1–2.5 mmol/L, pH 7.4). DPD was utilized at concentrations ranging from 1 to 100 μmol/L after being dissolved in dimethyl sulfoxide (DMSO) to create a stock solution (25 mmol/L). In some experiments, we used sodium-4-phenylbutyrate (4-PBA, stock solution: 50 mmol/L in DMSO, working concentration: 250 μmol/L) to inhibit ER stress.

After washing with Dulbecco’s phosphate buffered saline (DPBS), the cells were fixed in 4% paraformaldehyde for 20 min and rinsed with distilled water. Cells were stained with Alizarin Red S solution (2%, pH 4.2) for 10 min at room temperature. Excessive dye was removed by several washes in distilled water. To quantify AR staining, we added 100 μL of hexadecyl-pyridinium chloride solution (100 mmol/L) to each well and measured optical density (OD), using a microplate reader at 560 nm.

Cells grown on 96-well plates were washed twice with DPBS and decalcified with HCl (0.6 mol/L) for 30 min. The Ca content of the HCl supernatants was determined by the QuantiChrome Calcium Assay Kit. Following decalcification, cells were washed with DPBS and solubilized with a solution of NaOH (0.1 mol/L) and sodium dodecyl sulfate (0.1%), and the protein content of the samples was measured with the BCA protein assay kit. The Ca content of the cells was normalized to protein content and expressed as μg/mg protein.

Cells grown on 6-well plates were washed twice with DPBS and decalcified with 100 μL of EDTA (0.5 mol/L, pH 6.9) for 30 min. OCN content of the EDTA-solubilized ECM samples was quantified by an enzyme-linked immunosorbent assay according to the manufacturer’s protocol.

C57BL/6 mice (8–12-week-old male, n = 18) were exterminated by CO₂ inhalation and perfused with 5 mL of sterile DPBS. The entire aorta was harvested and cleaned under aseptic conditions, and cut into pieces. Aorta rings were maintained in control, high Pi + DPD (25 μmol/L), and high Pi+4-PBA (250 μmol/L) in DMEM supplemented with 10% FBS, antibiotic antimycotic solution, sodium pyruvate, L-glutamine, and 2.5 μg/mL Fungizone. After 7 days, the aorta pieces were washed in phosphate-balanced saline (PBS), opened longitudinally, and decalcified in 25 µL of 0.6 mmol/L HCl overnight. Ca content was determined by the QuantiChrom Ca-assay kit, as described previously.

Animal care and experimental procedures were performed following the institutional and national guidelines and were approved by the Institutional Ethics Committee of the University of Debrecen under registration number 10/2021/DEMÁB. Animal studies were reported in compliance with the ARRIVE guidelines. All the mice were housed in a temperature- (22°C) and light-controlled (12-h light/12-h dark) room, in cages with standard beddings and unlimited access to food and water. C57BL/6 mice (10 weeks old, male, n = 30) were randomly divided into three groups: control (Ctrl), CKD, and CKD + DPD (CKDD) (10 mice/group). CKD was induced by a two-phase diet, as described previously (Tani et al., 2017). In the first 6 weeks, the mice received a diet containing 0.2% adenine and 0.7% phosphate, followed by a diet containing 0.2% adenine and 1.8% phosphate for 3 weeks. Ctrl mice received a normal chow diet. DPD was suspended in 1% methylcellulose and administered orally at a dose of 15 mg/kg/day from week 7. Following the 9-week diet five mice/group were anesthetized with isoflurane and injected retro-orbitally with 2 nmol of OsteoSense dye that was dissolved in 100 µL of PBS. Twenty-4 hours later, mice were euthanized by CO₂ inhalation and blood was taken by heart puncture into K3-EDTA-containing tubes. Then mice were perfused with 5 mL of ice-cold PBS. Kidneys and aortas were isolated and analyzed immediately ex vivo by an IVIS Spectrum In Vivo Imaging System. We took kidney and aorta tissues out from the remaining 15 mice (5 mice/group), snap freeze them in liquid nitrogen and kept at −80°C for further analysis.

Serum urea, creatinine, phosphate and calcium levels were determined in mice by kinetic assays on a Cobas^® c501 instrument. K3-EDTA anticoagulated whole blood murine samples were analyzed by a Siemens Advia-2120i hematology analyzer with the 800 Mouse C57BL program of Multi-Species software. Hemoglobin concentration was measured by a cyanide-free colorimetric method. Hematocrit values were determined as a calculated parameter derived from red blood cell count (RBC in T/L) and mean cell volume (MCV in fL).

Total RNA was extracted from the kidney and aorta of C57BL/6 mice using Tri Reagent following the manufacturer’s protocol. RNA was reverse transcribed using a High-Capacity cDNA Reverse Transcription Kit. The qPCR reactions were carried out according to the protocol of the iTaq universal SYBR^® Green Supermix reagent, using primers listed in the “Resources table.” PCR was performed using a real-time PCR machine.

HAoSMCs were lysed in Laemmli lysis buffer. Proteins were resolved by SDS-PAGE (7.5% and 10%) and transferred onto nitrocellulose membranes. Western blotting was performed with the use of the primary antibodies listed in the “Resources table.” Following the primary antibody binding, membranes were incubated with horseradish peroxidase-linked rabbit and mouse IgG. Antigen-antibody complexes were visualized with the enhanced chemiluminescence system Clarity Western ECL. Chemiluminescent signals were detected conventionally on an X-ray film or digitally with the use of a C-Digit Blot Scanner. After detection, the membranes were stripped and reprobed for β-actin. Blots were quantified by using the built-in software on the C-Digit Blot Scanner.

To knockdown ATF4 gene expressions, we used Silencer^® select siRNA constructs targeting HIF-1α and ATF4. As a control, we used the negative control #1 construct. Lipofectamine^® RNAiMAX reagent was used to transfect HAoSMCs according to the manufacturer’s protocol.

Results are expressed as mean ± SD. At least three independent experiments were performed for all in vitro studies. Statistical analyses were performed with GraphPad Prism 8.0.1 software. Comparisons between more than two groups were carried out by a one-way ANOVA followed by Tukey’s multiple-comparisons test. To compare each of several treatment groups with a single control group, we performed a one-way ANOVA followed by Dunnett’s post hoc test. A value of p < 0.05 was considered significant.

Fifteen C57BL/6 mice (8–12 weeks old, male) were randomized into three groups (n = 5/group): control (Ctrl), CKD, and CKD treated with DPD (CKDD). CKD was induced with a 9-week-long, two-phase adenine- and high-phosphate-containing diet, as detailed in Figure 1A. DPD was administered orally at a dose of 15 mg/kg/day in the last 3 weeks of the experiment (Figure 1A). Ctrl mice received a normal chow diet. Hematological parameters, body weight, and kidney function were evaluated at the end of the experiment. Both CKD and CKDD mice lost approximately one-third of their initial weight during the experiment (Figure 1B). DPD completely corrected CKD-induced anemia revealed by similar hemoglobin, RBC count, and hematocrit values in Ctrl and CKDD mice (Figures 1C,E). Elevated plasma urea, creatinine, and phosphate levels indicated that the kidney function of the CKDD mice had declined to the same degree as that of the CKD mice (Figures 1F–I). CKD treatment did not change plasma calcium levels (Figure 1I).

CKD was associated with increased renal mRNA expression of specific hypoxia and ER stress markers, such as glucose transporter 1 (Glut1), ATF4, CHOP, and glucose-regulated protein 78 (GRP78) (Figure 1J). DPD treatment further exacerbated CKD-induced activation of HIF-1 target genes and ER stress markers in the kidneys (Figure 1J). In comparison to Ctrl, CKDD treatment triggered a 3-fold increase in Glut1, vascular endothelial growth factor A (VEGFA), and CHOP mRNA expressions in the aorta (Figure 1K). We observed marked upregulation of the protein expressions of ER stress markers ATF4 and CHOP in the kidneys of CKDD mice (Figure 1L).

Osteosense staining was performed to evaluate soft tissue calcification in Ctrl, CKD, and CKDD mice. CKD was associated with increased kidney and aorta calcification, which was further exacerbated by DPD treatment (Figures 2A–D). Aorta calcium measurement supported the pro-calcifying effect of DPD in CKD animals (Figure 2E). Calcification is a highly regulated process, similar to bone formation; therefore, next, we investigated the expression of osteo-/chondrogenic markers in kidney and aorta samples. Compared to Ctrl, Runt-related transcription factor 2 (Runx2), SRY-box transcription factor 9 (Sox9), bone morphogenetic protein 2 (BMP2), and Msh Homeobox 2 (Msx2) mRNA levels were higher in the kidneys of CKD mice. Furthermore, we noticed that CKDD mice had higher Sox9 and Msx2 mRNA levels than CKD animals had (Figure 2F). In the aorta, CKD triggered an increase in BMP2 mRNA expression compared to Ctrl, whereas CKDD induced marked elevations of Sox9, BMP2, and Msx2 mRNA levels (Figure 2G). Overall, these results show that DPD treatment induces hypoxia response and ER stress, increases osteo-/chondrogenic marker expressions, and promotes hydroxyapatite deposition in the kidney and aorta of CKD mice.

The stress signal network between hypoxia and ER stress is implicated in the progression of CKD; therefore, we further examined the effect of DPD on these pathways using an in vitro calcification model. Exposition of HAoSMCs to DPD (1–100 μmol/L) induced stabilization of HIF-1α and subsequent activation of the HIF-1 pathway, as revealed by a dose-dependent increase in Glut-1 protein expression (Figure 3A). We could not detect changes in HIF-1α mRNA levels in DPD-treated HAoSMCs, suggesting that DPD regulates HIF-1α in a post-transcriptional manner (Figure 3B). Hypoxia is a pathophysiological condition that induces ER stress through PERK; therefore, next, we investigated PERK activation in HAoSMCs in response to high Pi (2.5 mmol/L) with or without DPD (10 μmol/L). Pi-induced PERK phosphorylation was further exacerbated by DPD (Figure 3C). Furthermore, compared to control, the levels of phosphorylated eIF2α (P-eIF2α) were elevated by Pi and Pi + DPD (Figure 3C). The activation of the PERK pathway by Pi + DPD induced a massive upregulation of ATF4 mRNA and protein expressions (Figures 3D,E) as well as CHOP, and Grp78 (Figure 3F). Sustained ER stress can induce apoptosis, therefore next we investigated whether DPD influences cell viability. We performed MTT assay, and found that DPD (10 μmol/L) decreased cell viability in both normal and high Pi conditions (Figure 3G). Then, we addressed the pro-calcifying effect of DPD in HAoSMCs. As revealed by Alizarin red staining, DPD (10 μmol/L) largely intensified Pi-induced calcification (Figure 3H). The ECM of HAoSMCs treated with Pi + DPD had approximately 2.4 times more calcium deposition than the ECM of Pi-treated cells (Figure 3I). Moreover, OCN accumulation in the ECM of Pi + DPD-treated HAoSMCs was about 4-times higher compared to Pi-treated cells (Figure 3J).

After establishing that DPD induces ER stress and accelerates high Pi-induced calcification, we investigated whether ER stress plays a causative role in HAoSMC calcification triggered by Pi + DPD. First, we tested the effect of an ER stress inhibitor, 4-phenylbutyrate (4-PBA), on HAoSMC calcification. AR staining revealed that 4-PBA inhibited Pi + DPD-induced calcification of HAoSMCs (Figure 4A). Additionally, 4-PBA inhibited the accumulation of Ca and OCN in the ECM of Pi + DPD-treated HAoSMCs and attenuated ex vivo aorta calcification (Figures 4B–D). Furthermore, the knockdown of ATF4 by siRNA decreased Pi + DPD-induced calcification of HAoSMCs as evaluated by AR staining, as well as Ca and OCN measurements from the ECM (Figures 4E–H). These results show that DPD induces ER stress, particularly ATF4, which plays a crucial role in Pi + DPD-induced HAoSMC calcification.

After showing that both the HIF-1 pathway and ATF4 activation play essential roles in Pi + DPD-induced HAoSMC calcification, we wanted to understand whether there is a cross-communication between these two pathways. To this end, we applied HIF-1α targeted siRNA and examined the protein expression of HIF-1α and ATF4 in response to Pi (2.5 mmol/L), DPD (10 μmol/L), and Pi + DPD (Figure 5A). Western blot results revealed that the HIF-1α knock-down approach was successful and that in the absence of HIF-1α, DPD fails to upregulate ATF4 expression (Figure 5A). On the other hand, DPD induced HIF-1α expression regardless of the presence of ATF4 (Figure 5B). These results suggest a hierarchy between HIF-1α and ATF4 upon DPD treatment, in which HIF-1α is upstream of ATF4.