KAP1 modulates osteogenic differentiation via the ERK/Runx2 cascade in vascular smooth muscle cells

Background Osteoblast phenotypic transition in vascular smooth muscle cells (VSMCs) has been unveiled as a common cause of vascular calcification (VC). Krüppel-Associated Box (KRAB)-Associated Protein 1(KAP1) is a transcriptional corepressor that modulates various intracellular pathological processes from gene expression to DNA repair to signal transduction. However, the function and mechanism of KAP1 on the osteoblastic differentiation of VSMCs have not been evaluated yet. Methods and results We demonstrate that the expression of KAP1 in VSMCs is significantly enhanced in vivo and in vitro calcification models. Downregulating the expression of KAP1 suppresses the osteoblast phenotypic transition of VSMCs, which is indicated by a decrease in the expression of osteoblast marker collagenase type I (COL I) and an increase in the expression of VSMC marker α-smooth muscle actin (α-SMA). Conversely, exogenous overexpression of KAP1 could promote osteoblast phenotypic transition of VSMCs. Moreover, KAP1 upregulated the expression of RUNX family transcription factor 2 (Runx2), an inducer of osteoblast that positively regulates many osteoblast-related genes, such as COL I. Evaluation of the potential mechanism demonstrated that KAP1 promoted osteoblast phenotypic transition of VSMCs by activating the extracellular regulated protein kinases (ERK) signaling pathway, which could activate Runx2. In support of this finding, KAP1-induced cell osteoblast phenotypic transition is abolished by treatment with PD0325901, a specific ERK inhibitor. Conclusions The present study suggested that KAP1 participated in the osteoblast differentiation of VSMCs via the ERK/ Runx2 cascade and served as a potential diagnostics and therapeutics target for vascular calcification.


Introduction
Vascular calcification (VC), characterized by calcium and phosphate deposition in blood vessels, has been a life-threatening complication of atherosclerosis, diabetes, chronic kidney failure, and aging [1][2][3]. Studies have reported that VC is an active and tightly regulated process, sharing many similarities with physiology bone formation [4]. Vascular smooth muscle cells (VSMCs) are the major cell type of arterial media and essential regulators of vascular function. The transformation of VSMCs to an osteoblast-like phenotype is pivotal to the development of VC. This osteogenic transformation is demonstrated by the loss of SMC markers, including α-smooth muscle actin (α-SMA), and gain of osteochondrogenic markers, such as RUNX family transcription factor 2 (Runx2), collagenase type I (COL I), alkaline phosphatase (ALP) and osteopontin (OPN) [5]. In this regard, Runx2 plays a central role which is responsible for the expression of other bone-forming genes [6,7].
RAS-RAF-MEK-ERK is a major signaling pathway for cell proliferation, differentiation, apoptosis, and stress response under normal and pathological conditions [8][9][10]. Accumulating evidence has suggested that ERK signaling cascade is a key downstream pathway involved in cancer progression [11][12][13]. Likewise, the ERK pathway also participates in VC development and bone homeostasis [14]. A previous study has confirmed that inhibition of the ERK signaling pathway can reduce osteoblast differentiation of VSMCs [15].
Krüppel-associated box (KRAB)-associated protein1 (KAP1), also called triple motif protein 28 (TRIM28) or transcriptional intermediary factor 1 beta (TIF1-β), is an important cofactor of KRAB zinc finger proteins (KRAB-ZFPs) [16]. Its most described function is to target KRAB domain-containing-zinc finger proteins (KZFPs) to transposable elements (TE) by sequence-specific targeting and induce gene silencing [17][18][19][20][21][22]. In addition, transcription factors that lack a KRAB domain or any silencing complex also have been demonstrated to be regulated by KAP1 [23][24][25][26][27]. Evidence is increasing for the importance of KAP1 in a variety of intracellular processes. For instance, KAP1 directly bound to BCL2A1 and promoted BCL2A1 ubiquitination and subsequent proteasomal degradation to modulate cell death [28]. KAP1 enhances cell proliferation in glioblastoma by promoting autophagy [29]. In addition, knockdown of KAP1 abolished the effect of OxLDL on vascular endothelial dysfunction [30]. The positive expression of KAP1 is relative to phenotypic switching of human aortic smooth muscle cells (HASMCs) stimulated by platelet-derived growth factor subunit B homodimer (PDGF-BB). Knockdown of KAP1 prevented phenotypic switching of HASMCs induced by PDGF-BB [31]. Therefore, we speculate that KAP1 may also regulate the osteogenic differentiation in VSMCs after exposure to high phosphorus environment. Furthermore, KAP1 serves as a c-Raf binding partner, regulating the Ras-Raf-ERK signaling cascade to induce the growth of tumor [32]. Nevertheless, it is unclear whether KAP1 regulates the osteoblastic differentiation process of VSMCs through ERK pathway.
In the present study, we found that KAP1 expression was greatly increased in the aortas of calcified rats and VSMCs cultured in high phosphorus. In vitro experiment further demonstrated that KAP1 overexpression was promoted, whereas KAP1 deficiency attenuated high Pi-induced osteoblast phenotypic transition. Last, mechanistic analysis revealed that KAP1 induced VSMCs osteoblast differentiation via the ERK/Runx2 cascade.

Animals and groups
Twenty male Sprague-Dawley rats (aged 8-weeks) were purchased from Lab Animal Center of Hebei Medical University (Hebei Province, China). All the rats were kept in standard laboratory conditions (at 22 ± 2 °C in 55 to 60% relative humidity, under 12-h light/dark circadian cycle) and they had free access to standard food and tap water and food. After acclimatization, rats were randomly divided into two groups (n = 10): namely control group and VC group. Rat model of VC was constructed by 5/6 nephrectomy combined with a high phosphate diet protocol as previously described [33]. Briefly, first, rats were anesthetized and underwent a 2/3 left nephrectomy. One week later, the total right kidney was removed to achieve 5/6 nephrectomy. The control group underwent surgery to expose the kidneys, but no further procedures were performed. After surgery, the VC group was supplemented with calcitriol (600 ng/kg/ day), while the control group was fed with standard rodent chow for 20 weeks.

Cell culture
Isolation of VSMCs was performed from the aortas of Sprague-Dawley rats (8-weeks old) with the use of the explant method as described in a previous study [33]. After cultured in Dulbecco's Modified Eagle Medium containing 10% fetal bovine serum, cells were incubated at 37 °C in humidified 5% CO2. VSMCs were used from passages 3-6. For calcification, induction of VSMCs was conducted by meas of incubation in calcifying media containing 10 mM β-GP for 7 days. The medium was changed every 2 days. To explore whether the ERK pathway was involved in the VSMC phenotypic switching, VSMCs were treated with PD 0325901 (5 µM) for 48 h.

Cells transfection
The small hairpin RNA against rat KAP1 (shKAP1) and negative control shRNA (Scramble) were obtained from GenePharma Biotechnology Co. Ltd. (Shanghai, China). The target sequence of the shKAP1 was GCT CTC TAA GAA GCT GAT TTA. KAP1 expression plasmid (pcDNA3.1-KAP1(OE KAP1) was obtained from YouBio Biotechnology Co. Ltd. (Chongqing, China). Cells were seeded in 6-well plates for 12 h, then these corresponding plasmids were transfected into VSMCs for 48 h using Lipofectamine 3000™ following the manufacturer's instructions, and the cells were harvested for subsequent experiments.

Immunohistochemistry staining
Paraffin-embedded sections of thoracic aorta were permeabilized and dehydrated. Next, 3% H2O2 was added to block endogenous peroxidase. After being blocked by goat serum, the incubation of sections was performed with primary antibody overnight at 4 °C. The next day, after returning to room temperature, the sections were incubated with biotinbound secondary antibodies for 30 min at 37 °C, followed by staining with DAB. Finally, nuclei were counterstained with hematoxylin. The images were collected under an Olympus microscope.

Von kossa staining
Paraffin-embedded sections of thoracic aorta were permeabilized, dehydrated, incubated with a solution of 5% silver nitrate, and then irradiated with ultraviolet rays for 1 h. Then the samples were soaked in 5% sodium thiosulfate solution for 5 min, and in 0.1% hematoxylin-eosin for 2 min. Dehydrated samples were observed by light microscopy.

Alizarin red staining
After the medium had been removed, the cells were washed and then fixed in 4% paraformaldehyde at room temperature for 30 min, followed by incubation with 1% Alizarin red-Tris-HCL solution pH 4.2 (Solarbio Technology Co.; G1452) at 37 °C for 30 min. Then, calcified nodules were observed and photographed under an upright microscope.

Calcium content
Calcium content was determined with the use of a commercial kit (Beyotime, China), according to the manufacturer's protocol. Normalization of the results was performed to the total protein concentration (Solarbio, China).

ALP activity
Alkaline phosphatase (ALP) activity was measured using an ALP assay kit (Nanjing Jiancheng, China) according to the manufacturer's instructions. Calculated the results using curves generated from standard samples.

Immunofluorescence
VSMCs (50, 000 cells/well) were seeded on glass coverslips in 12-well plates. Following confluence, VSMCs were given different interventions. Next, the cells were washed, and fixed, followed by permeabilization with 0.3% Triton X-100 for 10 min, and blockage with goat serum at room temperature for 1 h. Then incubation of the cells was performed with primary antibodies at 4 ℃ overnight. On the second day, after washing, fluorescent secondary antibodies were used to incubate cells at 37 °C in the dark for 30 min. After washing, the cells were then stained with DAPI at Table 1 The sequences of  primers used for RT-qPCR  Gene  Forward primer  Reverse primer   KAP1  GGG CTA TGG CTT TGG GAC AGATG  GAT CCA GGC GTT CAA GGC TCAC  Runx2  CCG CAC GAC AAC CGC ACC AT  CGC TCC GGC CCA CAA ATC TC  COL I  GGT TTG GAG AGA GCA TGA CC  TTT GGG GAA ATG AGT TTG G  a-SMA  CAT CCG ACC TTG CTA ACG GA  GTC CAG AGC GAC ATA GCA CA  GAPDH  CAA GGT CAT CCA TGA CAA CTTTG  GTC CAC CAC CCT GTT GCT GTAG room temperature for 10 min. Finally, the fluorescence signal was detected under an Olympus microscope.

RNA isolation and quantitation
Total RNA was extracted from VSMCs with the use of TRIzol reagent, followed by amplification of cDNA with the use of a TB Green® Premix Ex Taq™ II Kit (Takara, China). Real-time qPCR was conducted to explore mRNA expression in triplicate and repeated in at least three separate experiments. The primer sequences with KAP1, Runx2, COL I, α-SMA, and GAPDH were provided by Sangon Biotech (Shanghai, China), which are shown in Table 1.

Western blotting analysis
Total cell protein was obtained with RIPA cell lysis buffer and collected by centrifugation. The protein concentrations in VSMCs were measured by BCA method (Solarbio, China). 30ug protein per lane was loaded on 10% SDS-PAGE, which was then transferred to PVDF membranes. After blockage in 5% skim milk at 37 °C for 1 h, incubation of the membranes was performed with specific primary antibodies overnight at 4 °C. The next day, after rinse in Tris buffered saline containing 1% Tween, the membranes were incubated with secondary antibodies at room temperature for 1 h. Antibody binding was measured using ECL detection reagent (Thermo, USA) and the intensity of protein bands was analyzed by using Image J software.

Statistical analysis
The analysis of results was performed by two-tailed Student's t-test and one-way analysis of variance (ANOVA). All statistical analyses were conducted with the use of software SPSS 17.0 and p < 0.05 was considered as statistically significant.

The expression of KAP1 was increased in calcified arteries and VSMCs.
A VC rat model was successfully induced, as evidenced by obvious von Kossa staining (Fig. 1A). The osteogenic marker COL I was distinctly upregulated in the calcified arteries compared to the control group (Fig. 1B). Meanwhile, the KAP1 was increased (Fig. 1B), whereas the VSMC marker α-SMA was markedly decreased in calcified arteries (Fig. 1B). VSMC calcification was evidenced by obvious Alizarin red S staining and increased calcium deposition in vitro ( Fig. 1C and 1D). The protein levels of KAP1 increased in calcified VSMCs, as indicated by western blot analysis (Fig. 1E), which is in parallel with the upregulation of COL I and downregulation of a-SMA protein expression (Fig. 1H). KAP1 mRNA was also upregulated (Fig. 1F), likewise, COL I mRNA was upregulated while a-SMA mRNA was downregulated (Fig. 1I) in calcified VSMCs.
Immunofluorescence also revealed an increase in KAP1 in high-Pi induced VSMCs (Fig. 1G, green). These results suggest that KAP1 may be participate in the calcification process through its connection with the osteoblast phenotypic transition of VSMCs.

KAP1 promoted osteogenic differentiation and calcification in VSMCs
Considering that the expression of KAP1 in VSMCs was significantly enhanced in vivo and in vitro calcification models, we then investigated whether up-regulated KAP1 under normal cultured conditions can reprogram VSMCs toward phenotype transforming and cause calcification. In order to assess the role of KAP1 in VSMCs of VC, we transfected the pcDNA3.1-KAP1 plasmid into normal cultured VSMCs. As shown in Fig. 2A, the pcDNA3.1-KAP1 plasmid upregulated the protein expression of KAP1 in VSMCs. Compared with the pcDNA3.1-GFP (Empty) group, up-regulation of KAP1 could increase the COL I protein expression and prevent the a-SMA protein expression in VSMCs under normal conditions ( Fig. 2A). Moreover, overexpression of KAP1 further aggravated ALP activity, calcium deposition and Alizarin red S staining of VSMCs (Fig. 2B-D).
Since elevated KAP1 may lead to calcification of VSMCs, we hypothesized that knockdown of KAP1 could alleviate calcification. To determine whether KAP1 is required for osteogenic differentiation and calcification of VSMCs, shRNA was used to knock down KAP1 in high Pi-treated VSMCs. As shown in Fig. 2E, KAP1 shRNA transfection significantly reduced the protein expression of KAP1 in VSMCs, which lead to a drop in COL I protein expression, Fig. 1 KAP1 expression was examined in calcified arteries and VSMCs n = 3 per group. A von Kossa staining of thoracic aortas from control rats and calcified rats. B Immunohistochemistry of KAP1, COL I, a-SMA expression in the thoracic aortas of control group and calcified group. C Alizarin red S staining showed calcified nodules in VSMCs cultured with normal medium (N) and high phosphate (β-GP) medium for 7 days. D Quantification of calcium content of VSMCs after culture with β-GP medium compared with N for 7 days. ***p < 0.001. E Real-time qPCR analysis of KAP1 expression in VSMCs cultured with β-GP for 7 days compared with N. ***p < 0.001. F Western blot and statistical analyses of KAP1 in VSMCs treated with N and β-GP medium for 7 days. ***p < 0.001. G Immunofluorescence showed the expression of KAP1 in normal medium and high-Pi medium cultured VSMCs. H Protein expression of COL I and a-SMA in VSMCs analyzed by western blot. ***p < 0.001. (I) mRNA levels of COL I and a-SMA estimated by quantitative RT-PCR. ***p < 0.001 ◂ while a-SMA increased. Similarly, the results of ALP activity, calcium deposition and Alizarin red S staining further verified the effect of KAP1 on osteoblast differentiation and calcification (Fig. 2F-H). These results suggest that osteochondrogenic markers expression and calcification were regulated by KAP1 in VSMCs.

ERK pathway was involved in osteoblast differentiation and calcification of VSMCs
In view of the ERK signaling pathway that has been reported to mediate the functional effect of osteogenesis [14,15,34], we questioned whether KAP1 affected ERK expression under osteogenic induction. First, we found that the ERK pathway was activated in calcified arteries and VSMCs stimulated by high phosphorus. As indicated in Fig. 3A, compared with paired control aortas, the intensity of p-ERK immunoreactivity was increased in calcified arteries. And again, western blot results showed p-ERK was increased in high Pi-induced VSMCs, but it did not alter ERK expression (Fig. 3B). To confirm the role of ERK in phenotypic switch regulation of VSMC, we applied the ERK pathway inhibitor PD0325901. As shown in Fig. 3C, PD0325901 significantly decreased p-ERK expression without any significant alteration in the expression of ERK, as compared with the control group with DMSO. Moreover, COL I was reduced by 24.6%, a-SMA was increased by 29.8% with the treatment of PD0325901 in VSMCs compared with the control group treated with DMSO. These findings demonstrated activation of the ERK pathway in calcified arteries and VSMCs.

The expression of Runx2 was controlled by KAP1 in VSMCs
Runx2 is an essential transcription factor which can regulate the downstream osteogenesis-related gene expression. Previous reports have shown that p-ERK could positively regulate Runx2 expression [35][36][37] and Runx2 could enhance CLO I expression [6,7]. Our results indicated that overexpression KAP1 upregulated the mRNA and protein levels of Runx2, whereas knockdown of KAP1 suppressed the expression of Runx2 ( Fig. 4A and C). Correlation analysis was performed to assess the relationship between the expression of KAP1 and Runx2. (Fig. 4B). The result showed that KAP1 was positively correlated with Runx2. Similarly, the results of the fluorescence intensity were consistent with the above indicators (Fig. 4D). These findings indicate that the expression of Runx2 is regulated by KAP1 in VSMCs.

KAP1 modulated osteogenic differentiation via the ERK-Runx2 cascade
To determine whether KAP1 affected ERK expression or activation, we transfected VSMCs with OE KAP1. The results indicated that KAP1 overexpression raised the protein level of p-ERK and the ERK expression remained unchanged (Fig. 5A). Moreover, no obvious alteration was witnessed in the expression of p38 with or without overexpression of KAP1, suggesting that p38 pathway was independent of KAP1 (Fig. 5A). In addition, correlation analysis showed KAP1 overexpression was significantly positively associated with p-ERK expression (Fig. 5B). These results revealed that during osteogenesis differentiation, ERK served as downstream of KAP1, whose regulatory function depended on the activity of p-ERK in high-Pi-induced VSMC.
To further verify the dependence of KAP1 on ERK signaling in VSMCs osteoblast differentiation, we transfected cells with KAP1 overexpression plasmid with or without ERK inhibitor PD0325901, then observing the alteration of Runx2, COL I and a-SMA of VSMCs. Western blot analysis suggested that KAP1 overexpression promoted Runx2 and COL I expression, and this effect was inhibited by blocking the ERK pathway using PD0325901 (Fig. 5C). Again, the alizarin red staining was consistent with these results (Fig. 5D). Taken together, these findings suggest that KAP1 accelerates osteoblast activity by modulating the ERK/ Runx2 cascade (Fig. 5E).

Discussion
In the current study, we elucidated the effect of KAP1 on the osteoblast differentiation and calcification of VSMCs. The findings suggested that KAP1 overexpression was enhanced, whereas KAP1 deficiency attenuated β-GP-induced osteoblast differentiation and calcification. Moreover, high-phosphate induced the activation of the ERK signaling pathway, and ERK signaling inhibition could alleviate osteoblast differentiation and calcification. Furthermore, KAP1 overexpression promoted the expression of Runx2 and COL I, which could be inhibited by blocking the ERK pathway. Thus, we summarized that KAP1 promotes the calcification Fig. 2 The regulation of KAP1 on osteoblast differentiation and calcification n = 3 per group. A Protein expression of KAP1, COL I and a-SMA in pcDNA3.1-GFP (Empty) and pcDNA3.1-KAP1 (OE KAP1) transfected VSMCs analyzed by western blot. ***p < 0.001 vs Empty group. B ALP activity, C calcium content and D Alizarin red S staining of VSMCs. ***p < 0.001. E Protein expression of KAP1, COL I and a-SMA in shRNA-negative control (Scramble) and shRNA-KAP1 (shKAP1) transfected VSMCs cultured in high-Pi medium. ***p < 0.001 vs Scramble group. F ALP activity, G calcium content and H Alizarin red S staining in VSMCs infected with shKAP1 compared with those infected with Scramble. ***p < 0.001 ◂ of VSMC in a high-phosphorus environment by activating the ERK pathway-mediated effect of promoting osteoblast differentiation. The current study highlights the key role of KAP1 in VC development.
KAP1, also known as TRIM28 or TIF1-β, is a large multi-domain protein that plays a multifaceted role in various organismal processes. Mouse embryos lacking KAP1 fail to develop into gastrulate, suggesting that KAP1 is vital for early post-implantation development [38]. The upregulation of KAP1 was detected in various types of tumors, including gastric, cervical, ovarian and breast cancers [39][40][41][42]; an increase in KAP1 expression enhanced the cell proliferation and migration, indicating that KAP1 promoted phenotypic transformation of tumor cells. Importantly, overexpression of KAP1 has been reported to induce the migration and invasion of HASMCs after exposure to PDGF-BB [31]. However, opposite conclusions were also reported. KAP1 and TRIM27 cooperatively maintain the contractile phenotype of human VSMCs, knockdown of KAP1 and TRIM27 cause phenotypic conversion of the contractile VSMCs to a synthetic phenotype [43], which suggested that the roles of KAP1 were divergent with cell-type-specific and the types of stimuli. In the current study, our results indicated KAP1 was significantly increased in calcified arteries and VSMCs. High KAP1 expression was correlated to the increase of ALP activity and calcium content, the formation of calcified nodules, and the expression of Runx2 and COL I. These findings suggest that high KAP1 expression is one of the biomarkers indicating the osteoblast differentiation and calcification of VSMCs and KAP1 might play a role in pro-vascular calcification in high-Pi induced VSMCs.
VC that is similar to bone formation is regulated by a number of signaling pathways, including ERK pathway [15,44]. Earlier research has found the activation of ERK signaling pathways promotes VC [45]. Moreover, it has reported that ERK signaling pathway is a crucial mechanism in osteoblast differentiation, and ERK inhibitors could weaken Runx2 expression and ALP activities [34]. Correspond to these findings, we demonstrated that ERK signaling pathway is highly activated in high-Pi induced VSMCs. Moreover, we also found that upregulation of KAP1 significantly increased the phosphorylation of ERK. Furthermore, we evaluated the effect of PD0325901, a specific ERK inhibitor, on KAP1-overexpressing VSMCs. The results revealed that treatment with 5 µM PD0325901 significantly inhibited the KAP1-overexpression-mediated increase in Runx2 and COL I expression, suggesting that ERK pathway may be the main downstream effector of KAP1 in promoting osteoblast differentiation and calcification of VSMCs.
VC that occurs in the end-stage of CKD mainly refers to the medial calcification, which is a cell-mediated process associated with mineral metabolic disorders and no efficient methods for pharmacotherapy [46]. Based on accumulating evidence, VC is an active and multifaceted biological process, including chronic inflammation, autophagy defects, endoplasmic reticulum (ER) stress, and osteoblast differentiation of VSMCs, suggesting that efforts at these cellular events should be made to develop new therapeutic strategies towards a more efficient targeting of VC [47].
Just as Shuai et al. were happyed to find a bone scaffold that favored to cell growth and bone formation [48], we were also trying to explore ways to inhibit the osteogenesis differentiation of VSMCs to restrain vascular calcification. This study indicates that KAP1 is highly expressed in calcified arteries and VSMCs and may play a role in pro-vascular calcification, which might be a potential novel target for the treatment of VC. Therefore, our findings would provide new insights into the pathophysiological mechanisms of VC, aiming to reduce the risk of cardiovascular complications. However, extensive in vivo and human researches are needed to confirm these findings and to pinpoint the exact molecular mechanism involved.

Conclusions
This study indicates that KAP1 is highly expressed in calcified arteries and VSMCs and may play a role in pro-vascular calcification in high-Pi induced VSMCs, and ERK pathway plays an important role in this process. Extensive in vivo and human researches are needed to confirm these findings and to pinpoint the exact molecular mechanism involved. were analyzed by western blot. ***p < 0.001. B Correlation analysis of protein levels for KAP1 and p-ERK (r = 0.8986, p < 0.05). C Protein expression of p-ERK, ERK, Runx2, COL I, and a-SMA in VSMCs were examined by western blotting. VSMCs treated with or without ERK inhibitor (PD0325901, 5 µM) for 24 h in normal medium. ***p < 0.001 vs OE KAP1 group. D Alizarin red S staining showed calcified nodules in VSMCs. (E) The proposed model that KAP1 induced osteogenic differentiation by ERK/Runx2 cascade in high-Pi cultured VSMCs ◂