A long non-coding RNA H19/microRNA-138/TLR3 network is involved in high phosphorus-mediated vascular calcification and chronic kidney disease

ABSTRACT

Vascular calcification, characterized by the accumulation of calcium-phosphate crystals in blood vessels, is a major cause of cardiovascular complications and chronic kidney disease (CKD)-related death. This work focuses on the molecules involved in high-phosphorus-mediated vascular calcification in CKD. A rat model of CKD was established by 5/6 nephrectomy, and the rats were given normal phosphorus diet (NPD) or high phosphorus diet (HPD). HPD decreased kidney function, increased the concentration of calcium ion and damaged vascular structure in the thoracic aorta of diseased rats. A high phosphorus condition enhanced calcium deposition in vascular smooth muscle cells (VSMCs). High phosphorus also increased the expression of RUNX2 whereas reduced the expression of α-SM actin in the aortic tissues and VSMCs. Long non-coding RNA (lncRNA) H19 was upregulated in the aortic tissues after HPD treatment. H19 bound to microRNA (miR)-138 to block its inhibitory effect on TLR3 mRNA and activated the NF-κB signaling pathway. Downregulation of H19 or TLR3 alleviated, whereas downregulation of miR-138 aggravated the calcification and vascular damage in model rats and VSMCs. In conclusion, this study demonstrates that the H19/miR-138/TLR3 axis is involved in high phosphorus-mediated vascular calcification in rats with CKD.

1. Introduction

Chronic kidney disease (CKD) is a worldwide public health issue which may lead to a considerable mortality rate due to the end-stage renal disease or its complications, including cardiovascular disease, cognitive dysfunction, inflammation, atherosclerosis syndrome, and malnutrition [1,2]. Vascular calcification, characterized by the accumulation of calcium-phosphate crystals in blood vessels, is a major cause of cardiovascular complications and CKD-related death [3]. Medial calcification reduces vessels elasticity, an effect known to be harmful, especially for large arteries such as the aorta, which consequently leads to heart dysfunction [4,5].

Vascular calcification is increasingly accepted as a cell-regulated biological process involving trans-differentiation of vascular smooth muscle cells (VSMCs) into osteo-/chondrocyte-like cells, which then triggers calcification in the vascular wall [6]. Similar to bone formation, the hallmark of the vascular calcification progress is accompanied with an upregulation of bone-related proteins such as runt-related transcription factor 2 (RUNX2) while a downregulation of contractile proteins such as alpha-smooth muscle actin (α-SM actin) [7,8]. Importantly, the phosphate homeostasis is impaired in the course of CKD [9]. During CKD progression, excessive phosphaturic hormones fails to control phosphate level and hyperphosphatemia develops [9]. Hyperphosphatemia is one of the major factors in CKD leading to vascular calcification and further development of CKD [10,11] and cardiovascular diseases [12]. A high-phosphorus condition is frequently applied to induce calcification models in CKD [13–16]. Identifying more molecules implicated in high-phosphorus-induced vascular calcification may provide novel ideas to the management of CKD-induced cardiovascular complications.

Although message RNAs (mRNAs) were initially largely focused, the important roles of non-coding RNAs (ncRNAs) in the process of vascular calcification have been increasingly concerned [17]. The ncRNAs are categorized into small or long ncRNAs (lncRNAs) depending on their length. LncRNAs are over 200 nucleotides (nts) in length and share many characteristics of mRNAs except for protein-coding functions [18]. MicroRNAs (miRNAs) are the most studied ncRNAs which play versatile roles in cellular functions due to their unique binding with the 3ʹuntranslated region (3ʹUTR) of target mRNAs and the subsequent gene degradation [19]. Importantly, dysregulation of either lncRNAs or miRNAs is crucial for vascular calcification [20,21]. A well-known RNA interaction model during post-transcriptional regulation is that some transcripts including lncRNAs may sequester a miRNA and then blocked its inhibitory function on other mRNA transcripts, termed competitive endogenous RNAs (ceRNAs) [22]. In this paper, by using integrated bioinformatic analyses, we predicted a potential lncRNA H19 (hereafter termed H19)/miR-138/TLR3 axis that is potentially implicated in vascular calcification. We aimed to explore the function of this axis in the vascular calcification in vivo using a rat model of CKD and in vitro using VSMCs in the setting of high phosphorus exposure.

2. Materials and methods

2.1. Rats with CKD and treatment

The study was ratified by the Animal Ethical Committee of China Medical University (Approval No. 15052111). All animal procedures were performed according to the Guide for the Care and Use of Laboratory Animals (NIH, Bethesda, Maryland, USA). Significant efforts were made to reduce the pain of conscious animals.

Male Sprague Dawley (SD) rats (200–250 g, 7 weeks old) were acquired from SLAC Laboratory Animal Co., Ltd. (Shanghai, China). The rats were assigned into following groups: normal diet group (Control group; rats were given normal diet; n = 5), high phosphorus group (HP group; rats were given high phosphorus diet; n = 5), sham operation + normal phosphorus diet group (Normal group; rats were subjected to sham operation and given normal phosphorus diet; n = 5), CKD + normal phosphorus diet group (NPD group; rats with CKD were given normal phosphorus diet; n = 5) and CKD + high phosphorus diet group (HPD group; rats with CKD were given high phosphorus diet; n = 35). The Normal group was set as the control for the NPD group to examine the effect of nephrectomy on rats. Meanwhile, the NPD group was set as the control for the HPD group to examine the function of HPD in rats with CKD, while the Control group was set as the control for the HP group to examine the role of high phosphorus alone in rats without CKD.

A total of 60 SD rats were subjected to 5/6 nephrectomy to induce CKD as previously reported [14]. The rats underwent two surgeries. All rats were fasted 12 h before surgery with free access to water and anaesthetized via intraperitoneal injection of 90 mg/kg 1% pentobarbital sodium. In the phase 1 surgery, the rats were fixed at a supine position and the bilateral kidney regions were exposed. An incision was produced on the abdominal cavity at 1.5 cm to the left-side ribs to expose the left-side kidney. Next, the perirenal fat sacs and arteries were isolated, and a microforceps was used to clip the arterial branches and ligate the artery at 1/3 to the upper and lower ends with only 1/3 blood supply maintained. One week later, the phase 2 surgery was performed and the right-side kidney was exposed in a similar manner. The perirenal fat sacs and arteries were isolated again. The complete right kidney tissue was resected, and then the muscularis and skin were sewed up. Among the 60 rats subjected to 5/6 nephrectomy, one rat died of massive hemorrhage during the phase 1 surgery. Two rats died of massive hemorrhage while one rat died of deep anesthesia during the phase 2 surgery. Two weeks after the surgeries, two rats died of abdominal infection, one died of suspected intestinal obstruction and one died of heart failure. As a consequence, a total of 52 rats with CKD were successfully induced. Five rats were allocated into the NPD group, 5 in HPD group. The remaining 42 rats with HPD were further allocated into six groups (n = 7 in each) and injected with lentiviral vector-based negative control (NC) of H19 (H19-NC), small interfering RNA (siRNA) of H19 (si-H19), si-H19 + miR-138 control, si-H19 + miR-138 inhibitor, miR-138 inhibitor + TLR3-NC or miR-138 inhibitor + si-TLR3. The lentiviral vectors (GeneCopoeia, Rockville, MD, USA) were injected at a dose of 1 × 10⁹ vector genome (vg). One week later, the lentiviral vectors were administrated once again. The rats were used for subsequent experiments after another 15 weeks. One rat in the H19-NC, si-H19 + miR-138 control, and miR-138 inhibitor + TLR3-NC groups died of infection, hunger strike and other reasons during the following 15 weeks. One rat was failed in modeling in the miR-138 inhibitor + TLR3-NC and miR-138 inhibitor + si-TLR3 groups, respectively. Five survived rats in each group were involved for subsequent experiments and data analyses.

For the sham operations, another five rats only had the left-side perirenal fat sacs and arteries isolated without ligation, and the right-side kidney was not resected to maintain the completeness of the kidney tissues. Rats in the Control, Normal and NPD groups were given normal phosphorus diet (0.09% Pi, 0.6% Ca and 20% protein concentration), while those in the HP (without any procedures) and HPD (with CKD) groups were given high phosphorus diet (1.2% Pi, 1.6% Ca and 20% protein concentration).

2.2. Examination of biomarkers

The serum biomarkers of rats were examined to identify whether the CKD in rats was successfully induced. In brief, serum samples were collected, in which the concentrations of urea, creatinine, calcium and phosphorus were examined using an automatic biochemical analyzer (LX20, Beckman Coulter, Inc., Chaska, MN, USA). Concentration of parathyroid hormone (PTH) in serum was examined using an enzyme-linked immunosorbent assay (ELISA) kit (60–2500, Rat Intact PTH ELISA Kit, Immutopics, Inc., San Clemente, CA, USA). All procedures were performed in strict accordance with the kit’s instructions. The optical density (OD) value at 450 nm was read using a multimode microplate reader (VICTOR X, PerkinElmer Inc., Waltham, MA, USA).

2.3. Examination of calcification in rat aorta

The rats were sacrificed through intraperitoneal injection of 150 mg/kg 1% pentobarbital sodium. After that, the rat chest was opened and the aorta was perfused under the pressure of 55 ± 5 mmHg. The right atrium was cut open to exclude blood and perfusion fluid. Next, the aorta was further perfused with 0.1 mol/L phosphate-buffered saline (PBS) containing 100 U/kg heparin for 5 min and 4% fresh paraformaldehyde at 4°C for 15 min. A section of thoracic aorta was dissected 1 cm away from the aortic arch, the perivascular tissue was removed, and the aortic tissues were washed and immersed in 4% fresh paraformaldehyde at 4°C for 3–4 h. After that, each thoracic aorta was transversely cut into 2–3 mm-thick 5–6 rings (2-3 mm thick) and embedded in paraffin for subsequent use [23].

The concentration of calcium ion in rat aorta was examined using the o-Cresolphthalein complexone method [24]. After animal death, the aorta tissues were digested in 1 mmol/L hydrochloric acid at 37°C for 24 h. The supernatant (decalcification solution) was collected and further centrifuged at 8,000 rpm for 10 min to collect the supernatant, in which the calcium concentration was examined using a calcium detection kit (MAK022, Sigma-Aldrich Chemical Company (St Louis, MO, USA). Every 20 mg tissue was added with 100–200 µL sample lysis buffer until fully lysed. The lysate was centrifuged at 10,000 g at 4°C for 3–5 min to collect supernatant. The supernatant was incubated with calcium working solution (1:1) in the dark at room temperature for 5–10 min, and then reacted with color development solution (1:1) for 10 min. The OD value at 575 nm was calculated to examine the calcium content in samples. The protein was extracted and quantified using the Lowry assay kit (Bio-Rad Laboratories, Inc., Hercules, CA, USA). The concentration of calcium relative to total protein was evaluated and expressed as calcium (μg)/protein (mg).

Calcification in rat aorta was detected by hematoxylin and eosin (HE) staining. In short, the tissues were fixed in 10% neutral formalin for 30 min and then cut into 5-μm sections. After rehydration, the sections were stained with hematoxylin solution (Sigma-Aldrich) for 5 min, and stained with eosin solution (Sigma-Aldrich) for 3 min. After that, the sections were sealed and observed under a fluorescence microscope (IX51, Olympus Optical Co., Ltd, Tokyo, Japan) [25].

2.4. LncRNA microarray analysis

The aortic vascular tissues of NPD and HPD rats were collected for lncRNA microarray analysis using GeneChip Rat Gene 2.0 ST Arrays (Affymetrix, Santa Clara, CA, USA) in strict accordance with the manufacturer’s instructions. Significant changes in gene expression profile were analyzed using the Affymetrix GeneChip Scanner 3000 7 G at a wavelength of 570 nm, and the statistical screening was performed using the Partek Genomic Suite 6.6 software (Partek Inc, MO, USA). The gene arrays of the core meta-sample set were normalized using the robust multiple-array average method. Gene expression was analyzed via the Fold Discovery Rate and a heatmap for differentially expressed genes was produced. Differentially expressed lncRNAs were screened using Foldchange > 2 and P < 0.01 as the thresholds.

2.5. Reverse transcription quantitative polymerase chain reaction (RT-qPCR)

The TRIzol reagent (Thermo Fisher Scientific Inc., Waltham, MA, USA) was used to collect total RNA from tissues and cells. After that, 2 μg RNA was reverse-transcribed to cDNA using an RNA First-Strand cDNA Synthesis Kit (K1621, Thermo Fisher Scientific). The reaction solution was collected for real-time qPCR. Using cDNA as the template, qPCR was conducted using an ABI7500 PCR kit (7500; Applied Biosystems, Foster City, CA, USA) in accordance with the TaqMan Gene Expression Assays protocol (4,331,182, Applied Biosystems). All primers were designed by Takara Biotechnology Ltd. (Dalian, China) and listed in Table 1, in which U6 was used as the internal control for miR-138 while GAPDH for other genes. Relative gene expression was examined by the 2^−ΔΔCt method [26]. ΔΔCt = ΔCt^{HPD group} – ΔCt^{NPD group}, in which ΔCt = Ct (target gene) – Ct (internal control).

Table 1.

Primer sequences for RT-qPCR.

Gene	Primer sequence (5’-3’)
miR-138	F: AGCTGGTGTTGTGAATCAGG
	R: GAACATGTCTGCGTATCTC
H19	F: TGTCAACAGGAAGGGAACGG
	R: CAGCTGCTTTACCTCGCTCT
TLR3	F: CCAGCCTTCAACGACTGAT
	R: CTGGGTTTGCGTGTTTCCAG
U6	F: CTCGCTTCGGCAGCACAT
	R: TTTGCGTGTCATCCTTGCG
GAPDH	F: GCATCTTCTTGTGCAGTGCC
	R: GATGGTGATGGGTTTCCCGT

Note: RT-qPCR, reverse transcription quantitative polymerase chain reaction; miR-138, microRNA-138, TLR3, toll like receptor 3; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; miR, microRNA; F: forward; R: reverse

2.6. Cell culture and identification

Another batch of male SD rats (80–100 g, 4 weeks old) acquired from SLAC were used to extract the VSMCs. The rats were euthanized by 1% pentobarbital sodium (150 mg/kg) and fixed. The chest was opened, and the rat aorta was taken out. The fat and connective tissues were instantly removed, and the aortic intima cells were collected and rinsed using Dulbecco’s modified Eagle’s medium (DMEM). The vessel was sliced into 1 mm³ size and cultured on culture dishes at 37°C with 5% CO₂ for 3 d. The tissues were digested in 0.25% trypsin, and the digestion was terminated by the addition of 10% fetal bovine serum (FBS), 100 U/mL penicillin and 100 μg/mL streptomycin in the DMEM. The cells were then resuspended and adjusted to 1 × 10⁴ cells/mL and sorted in 6-cm culture dishes. The culture medium was refreshed every 3 d and the cells started to subculture once reaching an 80% confluence. Cells at passage 6 to 12 were collected for further use [27].

The expression of VSMC-marker protein α-SM actin was examined by immunohistochemical (IHC) staining [28]). In brief, the paraffin-embedded aortic tissue sections were dewaxed at 67°C for 2 h. The sections were immersed in 100% alcohol for 5 min, 95% alcohol for 5 min, 85% alcohol for 3 min, and 75% alcohol for 2 min, and then rinsed with distilled water for 1 min. The tissue sections were soaked in 5% H₂O₂ for 30 min to diminish the activity of endogenous peroxidase, and then soaked in 0.01% Triton-100 and treated with 5% bovine serum albumin (BSA) for 20 min. After that, the cells were incubated with anti-α-SM actin (1:100, ab7817, Abcam Inc., Cambridge, MA, USA) at 4°C overnight, and then incubated with biotin-labeled goat anti-rabbit immunoglobulin G (IgG, 1:100, ab150113, Abcam) at 37°C for 20 min. A DAB kit (MK210, Takara) was used for color development. After that, the sections were counter-stained with hematoxylin at room temperature in the dark for 2 min and then viewed under an inverted microscope (IX51, Olympus) equipped with a photographing system to observe the positively stained cells. Positive staining rate in tissues was analyzed using the ImagePro Plus (NIH, USA).

2.7. Cell treatment

The VSMCs were sorted in 6-well plates at 1 × 10⁴ cells per well and allocated into Control and High Pi groups. Cells in the Control group were cultured in 10% FBS-supplemented low-glucose DMEM, and those in the High Pi group were further administrated with 10 mmol/L β-glycerophosphate. After 7 d, the concentrations of alkaline phosphatase (ALP) and calcium in each group of cells were examined. The si-H19, miR-138 inhibitor, si-TLR3 and the control vectors (100 pmol) were mixed with 250 μL serum-free Opti-MEM (Thermo Fisher Scientific) and incubated for 5 min, which were mixed with 5 mol/L Lipofectamine 2000 (Thermo Fisher Scientific) at room temperature for 20 min and transfected into cells as per the kit’s instructions. After 6 h of incubation at 37°C with 5% CO₂, the medium was replaced by complete medium (DMEM containing 10% FBS, 100 U/mL penicillin and 100 μg/mL streptomycin). Cells were used for further experiments at 48 h after transfection.

2.8. Examination of calcification in VSMCs

Calcium deposition in cells was examined by alizarin red staining [29]. In short, the cells were immobilized in 4% paraformaldehyde and the stained with 1% alizarin red solution (pH = 4.2, Leagene Biotechnology Co., Ltd., Beijing, China) at 30°C for 30 min. The cells were then rinsed with distilled water and captured under the optical microscope (Olympus).

The ALP expression in VSMCs was examined as an indicator of the osteogenic differentiation ability of cells. The VSMCs were cultured in 24-well plates at 4 × 10⁴ cells per well overnight. After three PBS washes, the ALP activity in cells was examined using an ALP detection kit (ab83369, Abcam). The ALP activity was normalized to total protein concentration and evaluated using the Bradford assay (5,000,201; Bio-Rad Laboratories) [30].

2.9. Subcellular localization of H19

The subcellular localization of H19 was first predicted on lncATLAS (http://lncatlas.crg.eu/) and further validated using nuclear-cytoplasmic RNA separation and fluorescence in situ hybridization (FISH) assays. The nuclear and cytoplasmic RNA was extracted using a cytoplasmic & nuclear RNA purification kit (21000; Norgen Biotek, St. Catharines, ON, Canada) according to the manufacturer’s instructions. In short, the cells were dissolved in Lysis Buffer J and centrifuged at 1,000 × g at 4°C for 10 min and loaded in spin column. Thereafter, the corresponding buffer and ethanol was added for RNA combination. A few moments later, the RNA impurities were washed away using hypotonic buffer, and Elution Buffer E was added to extract the RNA. The extracted RNA was further centrifuged at 15,000 × g for 20 min, after which the nuclear RNA (supernatant) and cytoplasmic RNA (sediment) were separated. The expression of H19 in nuclear RNA and cytoplasmic RNA was examined using RT-qPCR.

The FISH assay was conducted according to the protocol of a FISH Tag™ RNA Green Kit with Alexa Fluor™ 488 dye (F32952, Thermo Fisher Scientific) according to the manufacturer’s instructions. Once reaching an 80% confluence in 6-well plates, the cells were fixed, treated with proteinase K (2 μg/mL), glycine and phthalide reagent, and incubated with 250 μL pre-hybridizing solution at 42°C for 1 h. After that, the prehybridization solution was absorbed, and the cells were incubated with 250 μL probe (300 ng/mL)-contained hybridizing solution at 42°C overnight. Next, the nuclei were counter-stained with 4’, 6-diamidino-2-phenylindole (1:800) for 5 min. After that, the cell slides were quenched using anti-fluorescence quencher, and the cells were observed under the fluorescence microscope (IX51, Olympus) with five random fields included.

2.10. Dual-luciferase reporter gene assay

The putative binding sites (wild type, WT) between miR-138 and H19/TLR3 3ʹUTR were obtained from the Starbase system (http://starbase.sysu.edu.cn/), and the mutant type (MT) binding sites were designed. The WT and MT sites were synthesized by Sangon Biotech Co., Ltd. (Shanghai, China) and inserted into the pMIR-REPORT luciferase vectors (Thermo Fisher Scientific) to construct H19-WT/H19-MT and TLR3 3ʹUTR-WT/TLR3 3ʹUTR-MT luciferase vectors. These vectors were transfected with miR-138 inhibitor or miR-138 control into cells using the Lipofectamine 2000. After 24 h, the relative luciferase activity in cells was examined using a Dual-Luciferase Reporter Assay Kit (E1910, Promega Corp., Madison, WI, USA) according to the kit’s protocols [31].

2.11. Western blot analysis

Cells were lysed in radio-immunoprecipitation assay lysis buffer (Beyotime Biotechnology Co. Ltd., Shanghai, China) containing protease inhibitor for protein collection. The protein concentration was examined using a bicinchoninic acid kit (23225, Thermo Fisher Scientific). Next, 50 μg total protein was separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and loaded on polyvinylidene fluoride membranes. The membranes were blocked by 5% nonfat milk for 1 h and then hybridized with primary antibodies against RUNX2 (1:1,000, ab76956, Abcam), α-SM actin (1:500, ab7817, Abcam), NF-κB p65 (1:800, sc-8008, Santa Cruz Biotechnology, Inc, Santa Cruz, CA, USA) and GAPDH (1:2,000, ab125247, Abcam) at 4°C overnight, and with HRP-labeled goat anti-mouse IgG (1:2,000, ab6789, Abcam) at 37°C for 1 h. The blot bands were visualized using the enhanced chemiluminescence reagent (Beyotime) and analyzed using the Image Pro Plus 6.0 (Media Cybernetics Inc, Bethesda MD, USA).

2.12. Immunofluorescence staining

Cells were seeded into 24-well plates at 1 × 10⁴ cell/mL for 6 h until cell adherence. Thereafter, the cells were fixed with 4% paraformaldehyde at room temperature for 30 min, penetrated with 0.3% TritonX-100 (Sigma-Aldrich) for 10 min, and blocked with 3% bovine serum albumin for 30 min. After that, the cells were incubated with anti-NF-κB (1:200, ab207297, Abcam) at 4°C overnight, and then with goat anti-rabbit IgG (1:50, ab96899, Abcam) in the dark at room temperature for 2 h. The nuclei were stained with diluted DAPI (1:800), and the staining was observed under the fluorescence microscope (Olympus) [32].

2.13. Statistical analysis

All data were analyzed using SPSS22.0 (IBM Corp. Armonk, NY, USA). Data were shown as mean ± standard deviation (SD) from three independent experiments. Differences between two groups were analyzed using the unpaired t-test whereas those among multiple groups were analyzed by one- or two-way analysis of variance (ANOVA) followed by Tukey’s post-hoc test. p < 0.05 was regarded to show statistically significant difference.

3. Results

3.1. High phosphorus aggravates vascular calcification in rats with CKD

A rat model of CKD was constructed, and the model rats were given NPD or HPD. Compared to the sham-operated rats, the concentrations of serum urea and creatinine were increased in the diseased rats. In the CKD rats, HPD increased the concentration of phosphorus and PTH in the serum samples (Table 2). In addition, HPD led to a notable increase in the concentration of calcium ion in the thoracic aorta of rats compared to those in the Normal and NPD groups (Figure 1a). The subsequent HE staining showed that the thoracic aorta of the sham-operated rats had complete endothelial cells, consecutive spandex and regularly arranged VSMCs whose nuclei were stained in pink. In rats in the NPD group, the VSMCs were disorderedly arranged, and the vessel had a loose structure with a small number of black particles. For rats in the HPD group, there were a large number of black particles in the thoracic aorta. The VSMCs were disorderedly arranged, the spandex was broken and the vascular structure was severely disrupted (Figure 1b). Thereafter, total protein from the aorta tissues was extracted, in which the levels of calcification-related marker protein RUNX2 and VSMC marker protein α-SM actin were examined. Compared to the sham-operated rats (Normal group), the level of RUNX2 was increased whereas the level of α-SM actin was significantly reduced in the NPD-treated rats, and a more pronounced trend was found in the HPD-treated rats (Figure 1c), indicating that high phosphorus aggravates vascular calcification in rats with CKD. Subsequently, a lncRNA microarray analysis was performed to screen the differentially expressed lncRNAs in the aortic tissues between rats in the NPD and HPD groups. Importantly, HPD was found to significantly increase the H19 expression in rat aortic tissues (Figure 1d). RT-qPCR validated that the H19 expression was significantly increased in the vessel after HPD treatment. However, the HPD treatment did not affect the H19 expression in the non-CKD setting in control rats (Figure 1e). These results indicate that H19 plays an important role in HPD-mediated vascular calcification.

Table 2.

Serum biochemical parameters of normal and CKD rats fed an NPD or an HPD.

Biochemical Parameters	Normal (n = 5)	CKD
		NPD (n = 5)	HPD (n = 5)
Urea (mmol/L)	6.29 ± 0.91	28.42 ± 3.59*	30.16 ± 6.18*
Creatinine (μmmol/L)	66.85 ± 5.28	254.91 ± 30.12*	269.52 ± 26.14*
Calcium (mmol/L)	2.25 ± 0.21	2.34 ± 0.52	2.41 ± 0.36
Phosphorus (mmol/L)	1.49 ± 0.16	2.85 ± 0.94*	5.06 ± 0.61#
PTH (pg/mL)	25.16 ± 5.18	58.46 ± 9.52*	652.15 ± 72.55#

Note: CKD, chronic kidney disease; NPD, normal phosphorus diet group; HPD, high phosphorus diet; PTH, parathyroid hormone.

Figure 1.

High phosphorus aggravates vascular calcification in rats with CKD. A, concentration of calcium ion in rat aorta examined using the o-Cresolphthalein complexone method (*p < 0.05, one-way ANOVA); B, vascular structure in rat aorta observed by HE staining; C, protein levels of RUNX2 and α-SM actin in rat aortic vessel determined by western blot analysis (*p < 0.05, two-way ANOVA); D, differentially expressed lncRNAs between the aortic tissues from rats in the NPD and HPD groups screened using the microarray analysis; E, H19 expression in rat aortic vessels examined by RT-qPCR (*p < 0.05, one-way ANOVA (i) or unpaired t test (ii)). In the control group, rats were given normal diet; in the HP group, rats were given high phosphorus diet; in the normal group; rats were subjected to sham operation and given normal phosphorus diet; in the NPD group, rats were induced with CKD and given normal phosphorus diet; in the HPD group; rats were induced with CKD and given high phosphorus diet. Tissue samples were collected at 17 weeks after CKD induction/sham operation. In each group, n = 5.

3.2. High phosphorus aggravates calcification in VSMCs

VSMCs from rat aortic vessels were collected and cultured. Under the inverted microscope, each VSMC was in specific spindle shape with multiple cell processes and enriched cytoplasm, the kytoplasm was in high density and opaque, and nucleus was located in the center with multiple nucleoli, which are significant SMC characteristics (Figure 2a). The cells at passage 6 were collected for IHC staining, which confirmed positive staining of α-SM actin in cells (abundant brownish filaments of fibers in cytoplasm) (Figure 2b). The cells were cultured in a high-phosphorus condition, and then the calcification in VSMCs was evaluated by alizarin red staining. It was found that cells in the control group showed faint positive staining of alizarin red, whereas high phosphorus obviously enhanced the calcium deposition in cells (Figure 2c). In addition, a high phosphorus condition also induced ALP expression in the VSMCs (Figure 2d). Subsequently, the western blot analysis showed that the expression of α-SM actin in cells was significantly reduced whereas the RUNX2 expression was elevated in the VSMCs after high phosphorus treatment (Figure 2e). Moreover, in concert with the results in rat aortic vessels, a high-expression profile of H19 was detected in the VSMCs after high phosphorus treatment (figure 2f).

Figure 2.

High phosphorus aggravates calcification in VSMCs. A, morphology of VSMCs observed under a microscope; B, concentration and distribution of α-SM actin in VSMCs examined by IHC staining; C, calcium deposition in VSMCs examined by alizarin red staining; D, ALP activity in VSMCs determined by colorimetry (*p < 0.05, unpaired t test); E, mRNA expression of α-SM actin and RUNX2 in cells examined by western blot analysis (*p < 0.05, two-way ANOVA); F, expression of H19 in VSMCs examined by RT-qPCR (*p < 0.05, unpaired t test).

3.3. H19 serves as a sponge for miR-138

To examine the molecules mediated by H19, we first analyzed the subcellular localization of H19 in VSMCs. Data obtained from the lncATLAS system suggested that H19 was mainly located in the cytoplasm (Figure 3a). To validate this, the nuclear and cytoplasmic RNA of VSMC cells was separated, and an enrichment of H19 was confirmed in the cytoplasmic RNA (Figure 3b). A similar trend was confirmed by FISH which suggested that the positive staining of H19 was mainly distributed in the cytoplasm (Figure 3c). Subsequently, the candidate target miRNAs of H19 and their target mRNAs were predicted for a Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis. Most of the mRNAs were enriched on the NF-κB signaling pathway (Figure 3d). The western blot assay showed that the activity of the NF-κB signaling pathway was significantly enhanced in high phosphorus-treated rats and VSMCs (Figure 3e). Among the candidate target miRNAs of H19, eight miRNAs were concerned since their downstream mRNAs were suggested to be enriched on the NF-κB signaling pathway. Thereafter, the expression of these eight miRNAs in model rats was determined using RT-qPCR. Importantly, only miR-138 was downregulated in rats after HPD treatment, and downregulation of miR-138 was also observed in VSMCs cultured in the high-phosphorus condition (figure 3f). We then postulated that H19 possibly serves as a sponge for miR-138. The dual-luciferase reporter gene assay suggested that miR-138 inhibitor increased the activity of H19-WT luciferase vector in HEK-293 T cells, indicating a direct binding relationship between H19 and miR-138 (Figure 3g).

Figure 3.

H19 serves as a sponge for miR-138. A, subcellular localization of H19 predicted in the lncATLAS system; B, subcellular localization of H19 validated through a nuclear-cytoplasmic RNA separation assay (*p < 0.05, two-way ANOVA); C, subcellular localization of H19 validated through a FISH assay; D, a KEGG pathway enrichment analysis based on the candidate target mRNAs of the target miRNAs of H19; E, activity of the NF-κB signaling pathway in HPD-treated rats and in VSMCs examined by western blot analysis (*p < 0.05, one-way ANOVA); F, expression of eight candidate miRNAs in rat aorta (i) and expression of miR-138 in VSMCs (ii) examined by RT-qPCR (*p < 0.05, two-way ANOVA (i) or unpaired t test (ii)); G, binding relationship between H19 and miR-138 validated through a dual luciferase reporter gene assay (*p < 0.05, two-way ANOVA).

3.4. miR-138 targets TLR3 to mediate the NF-κB signaling pathway

The downstream mRNAs of miR-138 were predicted using several bioinformatic systems including StarBase (http://starbase.sysu.edu.cn/), TargetScan (http://www.targetscan.org/vert_72/), RNA22 (https://cm.jefferson.edu/rna22/), and miRTarBase (http://mirtarbase.cuhk.edu.cn/php/search.php) (Figure 4a). Consequently, 24 common genes were predicted, and eight of these mRNAs were analyzed to be enriched on the NF-κB signaling pathway (Figure 4b). Among the eight mRNAs, TLR3 was found to be increased in the aorta of HPD-treated rats and in high phosphorus-treated VSMCs (Figure 4c). Likewise, the binding relationship between miR-138 and TLR3 mRNA was validated by a luciferase assay. It was found that miR-138 inhibitor increased the activity of the TLR3-WT luciferase vector in cells (Figure 4d). Thereafter, si-H19, miR-138 inhibitor and si-TLR3 were administrated into the HPD-treated rats and in high phosphorus-exposed VSMCs. All transfections were successfully performed, and downregulation of H19 increased the expression of miR-138 while reduced the expression of TLR3 mRNA, whereas further miR-138 inhibition led to an increase in TLR3 expression (Figure 4e). The activity of the NF-κB signaling pathway in aortic tissues and cells was examined as well. It was found that the NF-κB activity in aortic tissues and VSMCs was reduced by si-H19, restored by miR-138 inhibitor, and further suppressed upon TLR3 inhibition (figure 4f). Moreover, the immunofluorescence staining also suggested that the nuclear translocation of NF-κB in cells was suppressed by si-H19 but increased by miR-138 inhibitor (Figure 4g). These results indicate that the H19/miR-138/TLR3 axis regulates the activity of the NF-κB pathway

Figure 4.

miR-138 targets TLR3 to mediate the NF-κB signaling pathway. A, a venn diagram for the intersected mRNAs of miR-138 predicted from four bioinformatic systems; B, a KEGG pathway enrichment analysis based on the 24 common mRNA outcomes; C, expression of eight candidate target mRNAs of miR-138 in aortic tissues and VSMCs examined by RT-qPCR (*p < 0.05, i: two-way ANOVA; ii: unpaired t test); D, binding relationship between miR-138 and TLR3 mRNA validated by a dual luciferase reporter gene assay (*p < 0.05, two-way ANOVA); E, expression of H19, miR-138 and TLR3 mRNA in aortic tissues and VSMCs after si-H19, miR-138 inhibitor and si-TLR3 transfections determined by RT-qPCR (*p < 0.05 vs. H19-NC; #p < 0.05 vs. si-H19 + miR-138 control; &p < 0.05 vs. miR-138 inhibitor + si-TLR3-NC, two-way ANOVA); F, activity of the NF-κB signaling pathway in aortic tissues and VSMCs after transfections examined by western blot analysis (*p < 0.05 vs. H19-NC; #p < 0.05 vs. si-H19 + miR-138 control; &p < 0.05 vs. miR-138 inhibitor + si-TLR3-NC, two-way ANOVA); G, nuclear translocation of NF-κB in VSMCs examined by immunofluorescence staining. In the control group, rats were given normal diet; in the HP group, rats were given high phosphorus diet; in the NPD group, rats were induced with CKD and given normal phosphorus diet; in the HPD group; rats were induced with CKD and given high phosphorus diet. Tissue samples were collected at 17 weeks after CKD induction (15 weeks after lentiviral vector transfection). In each group, n = 5.

3.5. The H19/miR-138/TLR3 axis mediates vascular calcification in model rats

The expression of miR-138 and TRL3 in rats was examined. The RT-qPCR results indicated that the expression of miR-138 was declined whereas the expression of TLR3 was increased in the HPD rats compared to the rats in the NPD and Normal groups. However, the high phosphorus treatment did not affect the expression of miR-138 and TLR3 in rats without CKD (Figure 5a). After that, the functions of the H19/miR-138/TLR3 axis on vascular calcification were examined. It was noteworthy that the concentration of calcium ion in thoracic aorta of HPD-treated rats was reduced after H19 downregulation but restored after further miR-138 inhibition. Importantly, the miR-138 inhibitor-mediated increase in calcium ion was blocked upon TLR3 silencing (Figure 5b). The HE staining showed that the vascular damages in rat aorta induced by high phosphorus were alleviated and the regular VSMC arrangement was recovered on H19 inhibition. However, the vascular damage was aggravated after miR-138 inhibition but relieved by si-TLR3 again (Figure 5c). The western blot analysis concerning the protein level of α-SM actin and RUNX2 was further performed. It was found that the protein level of RUNX2 were suppressed by si-H19 and si-TLR3 but enhanced by miR-138 inhibitor, whereas reverse trends were observed in terms of the protein level of α-SM actin (Figure 5d). These results validate that the H19/miR-138/TLR3 axis plays an important role in vascular calcification.

Figure 5.

The H19/miR-138/TLR3 axis mediates vascular calcification in model rats. A, expression of miR-128 and TLR3 in rat aortic vessels examined by RT-qPCR (*p < 0.05, two-way ANOVA); B, concentration of calcium ion in rat aorta examined using the o-Cresolphthalein complexone method (*p < 0.05 vs. H19-NC; #p < 0.05 vs. si-H19 + miR-138 control; &p < 0.05 vs. miR-138 inhibitor + si-TLR3-NC, one-way ANOVA); C, vascular structure in rat aorta observed by HE staining; D, protein levels of RUNX2 and α-SM actin in rat aortic vessel determined by western blot analysis (*p < 0.05 vs. H19-NC; #p < 0.05 vs. si-H19 + miR-138 control; &p < 0.05 vs. miR-138 inhibitor + si-TLR3-NC; two-way ANOVA). In the control group, rats were given normal diet; in the HP group, rats were given high phosphorus diet; in the Normal group; rats were subjected to sham operation and given normal phosphorus diet; in the NPD group, rats were induced with CKD and given normal phosphorus diet; in the HPD group; rats were induced with CKD and given high phosphorus diet. Tissue samples were collected at 17 weeks after CKD induction (15 weeks after lentiviral vector transfection). In each group, n = 5.

3.6. The H19/miR-138/TLR3 axis mediates calcification in VSMCs

The effects of the H19/miR-138/TLR3 axis on VSMCs were further analyzed. First, the alizarin red staining suggested the calcium deposition in VSMCs was reduced by si-H19, increased by further miR-138 inhibition, and reduced again upon TLR3 inhibition (Figure 6a). Similar trends were found concerning the ALP activity that the ALP concentration in VSMCs was reduced by si-H19 and si-TLR3 but increased by miR-138 inhibitor (Figure 6b). The expression of α-SM actin and RUNX2 in the VSMCs was examined as well. Downregulation of H19 or TLR3 led to an increase in the expression of α-SM actin whereas a decline in the expression of RUNX2 in cells, whereas miR-138 inhibition increased the expression of RUNX2 but reduced the expression of α-SM actin in the VSMCs (Figure 6c).

Figure 6.

The H19/miR-138/TLR3 axis mediates calcification in VSMCs. A, calcium deposition in VSMCs examined by alizarin red staining (*p < 0.05, two-way ANOVA); B, ALP activity in VSMCs determined by colorimetry (*p < 0.05 vs. H19-NC; #p < 0.05 vs. si-H19 + miR-138 control; &p < 0.05 vs. miR-138 inhibitor + si-TLR3-NC, one-way ANOVA); C, protein expression of α-SM actin and RUNX2 in cells examined by western blot analysis (*p < 0.05 vs. H19-NC; #p < 0.05 vs. si-H19 + miR-138 control; &p < 0.05 vs. miR-138 inhibitor + si-TLR3-NC, two-way ANOVA).

4. Discussion

Among the complications of CKD, cardiovascular diseases caused by dysregulated vascular calcification are the major contributors to CKD-related morbidity [33]. Current managements for vascular calcification are focusing on attenuating its causative factors [34]. During the course of CKD, hyperphosphatemia is maintained and serves as a serious risk factor for cardiovascular diseases, bone disease and mortality [12]. Therefore, reducing dietary phosphorus intake is highly recommended for patients and, controlling phosphorus in CKD is critical for a better outcome of patients [35,36]. In addition, exploring key molecules participating in high phosphorus-mediated vascular calcification may provide more options for the control of CKD complications. In the present work, we report that a H19/miR-138/TLR3 network is involved in high phosphorus-mediated vascular calcification in rats with CKD.

First, HPD aggravated the symptoms in model rats, as presented by increased concentration of calcium ion and disrupted vascular structure in the thoracic aorta of diseased rats. The high phosphorus treatment also increased calcium deposition in VSMCs. In experimental and clinical studies, high phosphorus has been correlated with multiple unfavorable outcomes of CKD-mineral bone disorders, including vascular calcification, bone mass loss, parathyroid hyperplasia, increased peripheral arterial stiffness and so forth [37–40]. A high inorganic phosphate level has been suggested to induce a loss in the expression of specific smooth muscle proteins in VSMCs and allow them to obtain an osteochondrogenic phenotype [41]. This was also validated by our present findings that high phosphorus induced an increase in ALP and RUNX2 expression in VSMCs whereas a decline in α-SM actin expression. Since the rat model was successfully constructed and HPD did show significant impact both in vivo and in vitro, we then focused on the molecules potentially involved.

Emerging evidence has shown the key functions of lncRNAs in vascular calcification. For instance, upregulation of lncRNA Lrrc75a-as1 was found to inhibit calcium accumulation in VSMCs [20]. Another lncRNA, ANCR, has been found to suppress β-glycerophosphate-induced osteoblastic differentiation and mineralization in VSMCs [42]. In this paper, the lncRNA microarray identified H19 as a significantly highly expressed in the aortic tissues in model rats given HPD. Although the direct role of H19 in the vascular calcification induced by high phosphorus has not been investigated, H19 has been suggested to be upregulated in calcific aortic valve disease and exert pro-mineralization function [43]. In this work, downregulation of H19 was induced in rats and in high-phosphorus-exposed VSMCs, after which the calcium deposition in aorta and VSMCs was reduced, along with a decline in the expression of RUNX2 and ALP whereas an upregulation in α-SM actin. Similarly, a recent research by Liu et al. suggested that H19 reduced RUNX2 expression and the p38 MAPK and ERK1/2 signal pathways to promote VSMC transition and calcification in vitro [30]. However, they did not elucidate the detailed mechanisms by which H19 mediates the RUNX2 changes and regulates the involved signaling pathways, and an in-vivo validation was not included. In our present work, we not only provided evidence that H19 is closely linked to the calcification of VSMCs in vitro, but also confirmed this important attribute of H19 in vivo that the H19 suppression reduced vascular calcification as well as the related molecules RUNX2 and α-SM actin in rat aorta. Moreover, we identified the related molecules via integrated bioinformatics analyses.

The candidate target miRNAs, and the target mRNAs of these miRNAs were obtained from several bioinformatics systems. The subsequent KEGG pathway enrichment analysis suggested that the related mRNAs are mainly enriched on the NF-κB signaling pathway. The NF-κB is critical for the process of vascular calcification and is abnormally activated in patients with CKD [44], and phosphate exposure has been evidenced to induce activation of the NF-κB pathway in VSMCs [45]. A NF-κB stimulator, serum- and glucocorticoid-inducible kinase 1, has been found to induce calcification in VSMCs as well [46]. We therefore considered that the NF-κB signaling pathway is implicated in the H19-mediated calcification events. After that, the bioinformatic analyses and experiments confirmed that miR-138, whose downstream mRNAs were enriched on the NF-κB signaling pathway, was a target miRNA of H19 and was downregulated in the tissues of HPD-treated rats and in high-phosphorus-exposed VSMCs. Although the correlation between miR-138 with VSMC differentiation, to the best of our knowledge, has not been concerned yet, miR-138 has been found to inhibit aortic valve calcification through suppressing osteoblastic differentiation of valvular interstitial cells [31]. Likewise, downregulation of miR-138 using oligonucleotide antimiR-138 has been reported to enhance the RUNX2 expression and the bone regeneration ability of bone marrow mesenchymal stem cell sheets [47]. Here, our experiments found that downregulation of miR-138 blocked the suppressive role of si-H19 on calcification in rat aortic tissues and VSMCs in high phosphorus conditions. Among the target mRNAs of miR-138, likewise, only TLR3 was highly expressed in the HPD-treated models and was enriched on the NF-κB signaling pathway. Importantly, on the basis of miR-138 inhibition, further downregulation of TLR3 reduced calcium accumulation in rat aortic tissues and VSMCs, and it recovered the expression of α-SM actin while reduced the expression of RUNX2. Interestingly, activation of TLR3 has been evidenced to trigger osteogenesis of aortic valve interstitial cells [48].

Collectively, this study reports a novel ceRNA network involved in vascular calcification in high-phosphorus conditions. During the process of high-phosphorus-mediated calcification in rats with CKD and in VSMCs, H19 is upregulated, and it binds to miR-138 and blocks its inhibitory effect on TLR3 mRNA, leading to further NF-κB activation (Figure 7). This study may offer novel insights into the management of high-phosphorus-mediated vascular calcification. However, due to the time and fund limits, the role of the H19/miR-138/TLR3 axis in the vascular calcification induced by other factors was not included in the present study. We would like to investigate this issue in our future works.

Figure 7.

A diagram for the molecular mechanism. High phosphorus aggravates calcification in rats with CKD and in VSMCs. H19 is upregulated during this process, which binds to miR-138 and blocks its inhibitory effect on TLR3 mRNA and activates the NF-κB signaling pathway.

Funding Statement

This work was supported by Youth Science Foundation of National Natural Science Foundation of China (NO. 81900642), the National Natural Science Foundation of China (NO. 82070763) and (NO. 81770766), China Postdoctoral Science Foundation (NO. 2021MD703904); Scientific Research Funding Project of Liaoning Provincial Department of Education “Seedling” Project of Young Scientific and Technological Talents (NO. QN2019006) and Youth Backbone Project of China Medical University (NO. 1210519033)

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.