Abstract

   Vascular smooth muscle cell (VSMC) proliferation is a hallmark of
   neointimal hyperplasia (NIH) in atherosclerosis and restenosis
   post-balloon angioplasty and stent insertion. Although numerous
   cytotoxic and cytostatic therapeutics have been developed to reduce
   NIH, it is improbable that a multifactorial disease can be successfully
   treated by focusing on a preconceived hypothesis. We, therefore, aimed
   to identify key molecules involved in NIH via a hypothesis-free
   approach. We analyzed four datasets ([41]GSE28829, [42]GSE43292,
   [43]GSE100927, and [44]GSE120521), evaluated differentially expressed
   genes (DEGs) in wire-injured femoral arteries of mice, and determined
   their association with VSMC proliferation in vitro. Moreover, we
   performed RNA sequencing on platelet-derived growth factor
   (PDGF)-stimulated human VSMCs (hVSMCs) post-phosphoenolpyruvate
   carboxykinase 2 (PCK2) knockdown and investigated pathways associated
   with PCK2. Finally, we assessed NIH formation in Pck2 knockout (KO)
   mice by wire injury and identified PCK2 expression in human femoral
   artery atheroma. Among six DEGs, only PCK2 and RGS1 showed identical
   expression patterns between wire-injured femoral arteries of mice and
   gene expression datasets. PDGF-induced VSMC proliferation was
   attenuated when hVSMCs were transfected with PCK2 siRNA. RNA sequencing
   of PCK2 siRNA-treated hVSMCs revealed the involvement of the
   Akt-FoxO-PCK2 pathway in VSMC proliferation via Akt2, Akt3, FoxO1, and
   FoxO3. Additionally, NIH was attenuated in the wire-injured femoral
   artery of Pck2-KO mice and PCK2 was expressed in human femoral
   atheroma. PCK2 regulates VSMC proliferation in response to vascular
   injury via the Akt-FoxO-PCK2 pathway. Targeting PCK2, a downstream
   signaling mediator of VSMC proliferation, may be a novel therapeutic
   approach to modulate VSMC proliferation in atherosclerosis.

   Keywords: vascular smooth muscle cell proliferation, neointimal
   hyperplasia, atherosclerosis, PCK2, Akt-FoxO-PCK2 pathway

Introduction

   Atherosclerosis is the leading cause of some of the deadliest diseases,
   such as myocardial infarction and stroke, worldwide [45]^1.
   Pathologically, it involves multiple processes, such as oxidative
   stress, endothelial dysfunction, lipid infiltration, chronic
   inflammation, cell proliferation and migration [46]^2. Several studies
   have suggested that aberrant vascular smooth muscle cells (VSMCs)
   proliferation and neointima formation dominate the development of
   atherosclerotic lesions [47]^3. Blood vessel walls are majorly composed
   of VSMCs, which are crucial in maintaining the normal physiological
   function of blood vessels [48]^4. Notably, phenotype switching from
   contractile VSMCs to synthetic VSMCs increases the proliferation and
   migration of VSMCs and the expression of the extracellular matrix.
   Moreover, this is an major step in the growth of intimal hyperplasia
   [49]^5^, [50]^6.

   Understanding the molecular and cellular mechanisms underlying VSMC
   proliferation in atherosclerosis is a cornerstone for the development
   of antiproliferative therapeutic approaches. Studies have investigated
   various cell cycle-related proteins and pathways in association with
   VSMC proliferation. For example, overexpression of p21, a
   cyclin-dependent kinase inhibitor 1, using an adenoviral vector in
   balloon-injured porcine arteries limited the development of intimal
   hyperplasia [51]^7. Moreover, inhibition of cyclin-dependent kinase 2
   and repression of cyclin A gene transcription, via the induction of
   p27, inhibited VSMC proliferation in rat carotid artery after balloon
   angioplasty [52]^8. Moreover, mounting evidence suggests that noncoding
   RNA plays critical roles in regulating VSMC proliferation. For
   instance, long intergenic noncoding RNA-p21 represses VSMC
   proliferation and induces cell apoptosis by enhancing p53 activity
   [53]^9. Additionally, miR-378a-5p promoted VSMC proliferation by
   repressing the CDK1 gene expression [54]^10.

   In the clinical scenario, two drugs are widely used to treat
   atherosclerosis. They are delivered directly to the blood vessel wall
   in the form of coated balloon catheters and stents. One of these drugs,
   rapamycin, is a macrocyclic triene antibiotic that is the product of a
   natural fermentation process. Rapamycin interferes with mTOR and its
   downstream signaling cascades, reducing protein synthesis and causing
   concomitant cell cycle arrest [55]^11. In several clinical trials,
   implantation of rapamycin-coated stents on the coronary artery caused a
   considerable decrease in the restenosis rate and cardiovascular events
   [56]^12. The second drug, paclitaxel, is a potent antimicrotubule agent
   that promotes the development of stable dysfunctional microtubules that
   disrupt cellular processes [57]^13. Paclitaxel-coated balloons and
   stents are superior in targeting lesion revascularization to
   non-paclitaxel-coated devices and, therefore, have been widely accepted
   in clinical practice [58]^14. These advances in target lesion
   revascularization in peripheral arterial disease have decreased major
   adverse limb events, such as amputation and development of acute limb
   ischemia or critical limb ischemia, post-revascularization. However,
   there are still unmet needs, such as in-stent restenosis, and
   repetitive interventions are required to maintain patency [59]^15. It
   is crucial to emphasize that it is improbable that a multifactorial
   disease, such as atherosclerosis, can be successfully treated by
   focusing on a preconceived hypothesis of cytotoxic or cytostatic
   mechanisms. Therefore, in this study, we use a hypothesis-free approach
   to identify a key regulator of VSMC proliferation in response to
   vascular injury, analyze its pathway, and verify its function in vitro
   and in vivo.

Materials and Methods

Data acquisition

   We obtained the gene expression profiles for carotid atheroma from the
   Gene Expression Omnibus database (GEO;
   [60]http://www.ncbi.nlm.nih.gov/geo/). The accession numbers were
   [61]GSE43292[62]^16, [63]GSE28829 [64]^17, [65]GSE100927 [66]^18, and
   [67]GSE120521 [68]^19. Although the [69]GSE120521 dataset included gene
   expression profiles obtained by high throughput sequencing, the
   remaining three datasets included gene expression profiles obtained by
   array sequencing. Two of these datasets, [70]GSE43292 and
   [71]GSE100927, contained gene expression data from disease (carotid
   atheroma or atherosclerotic carotid artery) groups and/or normal
   (intact carotid artery) groups. Additionally, [72]GSE28829 included
   gene expression data associated with early and advanced carotid
   atheroma plaques, whereas [73]GSE120521 included gene expression data
   associated with stable and unstable carotid atheroma plaques. Detailed
   characteristics of these datasets are presented in [74]Supplementary
   Table 1.

Data preprocessing

   There is an inevitable heterogeneity in the data published in the GEO
   database. We performed data preprocessing using the “rma,” “voom,” and
   “loess normalization” methods provided by the “limma” package of R
   software (v 3.42.2) to minimize heterogeneity [75]^20. We converted the
   probe ID into a gene symbol and used it, and when several probes were
   matched with one gene symbol, the average level of the probe was used
   as the expression value of the corresponding gene. In addition, the
   absence of a gene symbol was excluded from this study.

Identification of differentially expressed genes

   To identify the differentially expressed genes (DEGs) between carotid
   atheroma and normal tissues in the GEO dataset, we conducted linear
   regression analysis using the “limma” package. We considered genes with
   an adjusted P value < 0.05 to have a significant differential
   expression between the two groups. The false discovery rate (FDR)
   method was used to calculate the adjusted P value. The log[2] fold
   change > 0 was the threshold for identifying up-regulated genes, and a
   log[2] fold change < 0 was the threshold for identifying down-regulated
   genes. To screen for the intersectional genes significantly expressed
   in each cohort, we used the R package “venn” (v1.10) to plot Venn
   diagrams.

Identification of potential PCK2-related genes and signaling pathways

   We performed DEG analysis using RNA-seq data from cells harboring
   phosphoenolpyruvate carboxykinase 2 (PCK2) knockdown to identify genes
   potentially associated with PCK2. We performed DEG analysis using the
   “limma” package, and the criteria for PCK2 associated DEG were defined
   as adjusted P < 0.05 and -2 > log[2] fold change > 2. To identify the
   potential signaling pathways associated with PCK2, we performed a Kyoto
   Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis
   using the R package "clusterProfiler" (v 3.14.3) [76]^21. The cut-off
   values for KEGG analysis were set according to parameters, and terms
   with ≥ 5 associated genes and having an adjusted p-value < 0.05 were
   included for subsequent analyses.

Animal experimentation and ethics statement

   All experiments involving animals complied with the guidelines provided
   by the Guide for the Care and Use of Laboratory Animals published by
   the US National Institute of Health (NIH Publication No. 85-23, 2011
   revision). All animal experiment protocols were permissioned by the
   Pusan National University Institutional Animal Care and Use Committee
   of the College of Medicine (PNU-2020-2680). We performed genotyping,
   including that of Pck2-deficient mice, by PCR, using the protocol
   provided by the Jackson Laboratory (Harlan Nossan, ITA). We purchased
   C57BL/6J wild-type control mice (WT) from the Jackson Laboratory. All
   animals were housed in an air-conditioned room at 22-25 °C and
   maintained under a 12 h light/dark cycle. Food and water were provided
   ad libitum.

   All human experiments in this study complied with the Declaration of
   Helsinki and were approved by the locally appointed ethics committee of
   the Gachon University Gil Medical Center (GFIRB2020-085). Informed
   consent was obtained from all the study participants.

Vascular injury models and blood flow measurement

   We anesthetized the WT and Pck2-deficient male mice (seven weeks old)
   via an intraperitoneal injection of chloral hydrate (450 mg/kg).
   Efficiency of an anesthetic dose was confirmed by the response to a toe
   pinch. Subsequently, we injured them in the right femoral artery using
   a 0.26 mm diameter angioplasty guidewire with hydrophilic coating
   (ASAHI RG3, Asahi Intecc, Japan) under aseptic conditions, as
   previously described [77]^22. The wire-injured femoral arteries were
   harvested from mice euthanized by CO[2] insufflation four weeks after
   the injury. Isolated femoral arteries were fixed overnight in 4%
   paraformaldehyde at 4 °C, and their paraffin-embedded cross sections (4
   μm) were prepared. These tissue sections were stained with hematoxylin
   and eosin (H&E) and specific antibodies for immunohistological
   analysis. The neointimal volume was measured using an Image J, version
   1.53m, program (NIH Image, Bethesda, MD). Blood flow in the femoral
   artery was measured using a laser Doppler perfusion imaging (LDPI)
   analyzer (Moor Instruments, Devon, UK) at 0 (before wire-injury), 1, 2,
   3, and 4 weeks after the femoral artery injury. Blood flow was
   quantified using pixel values of LDPI color.

Chemical reagents and antibodies

   Antibodies against PCK2 (Cat #. NBP2-75610), KIAA1671 (Cat #.
   NBP1-91085), FHL1 (Cat #. NB100-1460), and LAMA5 (Cat #. NB300-144)
   were purchased from Novus Biologicals (Littleton, CO, USA), and those
   against RGS1 (Cat #. MBS851803) and AMPD3 (Cat #. MBS766604) were
   purchased from MyBioSource (San Diego, CA, USA). Platelet-derived
   growth factor (PDGF) and α-smooth muscle actin (α-SMA) (Cat #. A2547)
   were purchased from Millipore-Sigma (St. Louis, MO, USA). Antibodies
   against Ki67 (Cat #. SC-7846), PCNA (Cat #. SC-56), and β-actin (Cat #.
   SC-47778) were purchased from Santa Cruz Biotechnology, Inc. (Beverly,
   MA, USA). The PI3K inhibitor (LY294002; LY) and the Akt inhibitor (Akt
   inhibitor IV; AI) were purchased from Calbiochem (La Jolla, CA, USA).
   Antibodies against Akt1 (Cat #. 75692S), Akt2 (Cat #. 5239S), Akt3 (Cat
   #. 14982S), FoxO1 (Cat #. 2880S), FoxO3 (Cat #. 12829S), FoxO4 (Cat #.
   9472S), and MitoTracker (Cat #. 8778S) were purchased from Cell
   Signaling Technology (Beverly, MA, USA). Phospho-FoxO1 (Cat #.
   ab131339), phospho-FoxO3 (Cat #. ab47285), phospho-FoxO4 (Cat #.
   ab47278), COXIV (Cat #. ab33985), and β-tubulin (Cat #. ab131205)
   antibodies were purchased from Abcam (Cambridge, UK). The PCK2
   inhibitor PEPCKi was obtained from Axon Medchem (Reston, VA, USA).
   Horseradish peroxidase (HRP)-conjugated IgG antibody (Santa Cruz
   Biotechnology, Inc.) was used as the secondary antibody.

Cell culture

   Human aortic vascular smooth muscle cells (hVSMCs) were obtained from
   the American Tissue Culture Collection (Manassas, VA, USA). We cultured
   the hVSMCs in culture dishes containing a Medium-231 (M231500, Gibco
   BRL, Grand Island, NY, USA), 10% fetal bovine serum, 1%
   antibiotic-antimycotic solution (Gibco BRL) and smooth muscle growth
   supplement (S00725, Gibco BRL). Cells were incubated at 37 °C in a
   humidified atmosphere containing 5% CO[2].

Western blot analysis

   We prepared hVSMC lysates in an ice-cold RIPA lysis buffer (Thermo
   Fisher Scientific, Rockford, IL, USA). The cytosolic and mitochondrial
   proteins of hVSMCs were extracted with Mitochondria/Cytosol
   Fractionation Kit (Cat #. ab65320, Abcam) according to the
   manufacturer's protocol. Subsequently, equal amounts (30 μg) of protein
   were separated by 8-10% SDS-PAGE under reducing conditions and then
   transferred to nitrocellulose membranes (Amersham Pharmacia Biotech,
   Piscataway, NJ, USA). The protein transferred membrane was blocked with
   5% skimmed milk in Tris-buffered saline-Tween-20 (TBST) for 2 h at room
   temperature. Thereafter, they were incubated overnight with primary
   antibodies (1:1000) diluted in 5% skim milk at 4 °C. The following day,
   the blots were washed with TBST and were incubated with HRP-conjugated
   secondary antibody (1:3000) for 2 h at room temperature. Subsequently,
   blots were developed using the chemiluminescent ECL detection reagent
   (Thermo Fisher Scientific). An anti-β-actin antibody was used as an
   internal control. Protein bands were quantified using the UN-SCAN-IT
   GEL 7.1 program (Silk Scientific, UT, USA), and protein expression
   levels were calculated in relation to the β-actin protein expression.

Reverse transcription (RT)-PCR analysis

   Total RNA was isolated from the cells using the TRIzol reagent
   (Invitrogen, Carlsbad, NY, USA), according to the manufacturer's
   protocol. The isolated RNA (1 μg) by applying the ImProm-II reverse
   transcription system (Promega, Madison, WI, USA) was reverse
   transcribed into cDNA. Subsequently, cDNA amplification was performed
   using PCK2-specific primers (forward, 5′-GGG TGC TAG ACT GGA TCT GC-3′
   and reverse, 5′-CTG GTT GAC CTG CTC TGT CA-3′). Equal amounts (5 μL) of
   PCR products were identified on a 0.8% agarose gel containing ethidium
   bromide. PCR products were quantified using the UN-SCAN-IT GEL 7.1
   program, and the mRNA expression of PKC2 was calculated in relation to
   that of GAPDH.

Small interfering RNA (siRNA) preparation and transfection

   We designed and synthesized siRNAs for PCK2, RGS1, Akt1, Akt2, Akt3,
   FoxO1, and FoxO3 and scrambled siRNA as a negative control (Cat no.
   SN-1003) using an AccuTarget siRNA construction kit (Bioneer, Daejeon,
   Korea). All siRNAs were transfected through the pores of the cell
   membrane using Lipofectamine 2000 (Invitrogen) according to the
   manufacturer's protocol.

RNA extraction, library construction, and sequencing

   Total RNA (0.5 μg) was extracted from the transfected hVSMCs using the
   TRIzol reagent kit (Invitrogen), and the integrity of the RNA was
   assessed using the TapStation RNA screen tape. And then, RNA libraries
   were prepared using the Illumina TruSeq Stranded Total RNA Library Prep
   Gold Kit (Illumina, Inc., San Diego, CA, USA). The cleaved RNA fragment
   was copied into first-strand cDNA using random primers (Invitrogen) and
   reverse transcriptase (Invitrogen, Carlsbad, CA, USA). Finally, the
   cDNA library that passed the quality check was sequenced using the
   Illumina NovaSeq platform (Illumina, Inc., San Diego, CA, USA). All
   kits were used according to the manufacturer's protocol.

RNA-seq analysis

   We preprocessed the raw reads obtained from the sequencer to remove low
   quality reads and adapter sequences before performing the RNA-seq
   analysis. We aligned the processed reads to Homo sapiens (hg38) genome
   sequence using HISAT v2.1.0(1). This reference genome and the relevant
   annotation data were downloaded from the University of California Santa
   Cruz table browser ([78]http://genome.uscs.edu). After alignment,
   StringTie v1.3.4 was used to estimate transcript abundance in each
   sample. The estimated transcript abundance table was used for
   subsequent analysis.

MTT assay

   Proliferation rates of hVSMCs were analyzed using the MTT assay.
   Briefly, 1 × 10^5 cells were cultured in MTT solution (EZ-Cytox, Daeil
   Laboratories, Seoul, Republic of Korea)-treated medium according to the
   manufacturer's protocol. Cells were incubated at 37 °C in a humidified
   atmosphere containing 5% CO[2]. Optical density of the formazan
   crystals was measured at 450 nm using a microplate reader (TECAN
   SUNRISE, Mannedorf, Switzerland). Cells were counted using a
   hemocytometer. Relative proliferation rates were analyzed by comparing
   PDGF-treated cells with control cells.

Immunofluorescence analysis

   We incubated serial paraffin-sectioned (4 μm) femoral arteries with
   mouse anti-α-SMA (1:400) and either rabbit-anti PCK2 (1:200),
   rabbit-anti RGS1 (1:200), rabbit-anti AMPD3 (1:200), rabbit-anti
   KIAA1671 (1:200), goat-anti FHL1 (1:200), or rabbit-anti LAMA5 (1:200)
   antibodies. hVSMCs, plated on glass coverslips, were fixed with 4%
   paraformaldehyde, and nonspecific binding sites were blocked with 1%
   bovine serum albumin. The fixed cells were then incubated with
   MitoTracker, rabbit anti-PCK2 (1:400), and mouse anti-β-tubulin (1:400)
   antibodies. Goat anti-mouse Alexa488-conjugated IgG and Goat anti-mouse
   Alexa594-conjugated IgG or Donkey anti-goat Alexa 594-conjugated IgG
   (Abcam) were used to label the immunofluorescence signals for α-SMA,
   PCK2, RGS1, AMPD3, KIAA1671, FHL1, and LAMA5, respectively. After
   nuclei were stained with DAPI at 0.1 μg/mL, slides were mounted on a
   Vectashield, and fluorescence images were visualized using a scanning
   confocal microscope (LSM 510, Carl Zeiss, Oberkochen, Germany).

Human femoral atheroma sampling

   Five patients who underwent common femoral endarterectomy at Gachon
   University Gil Medical Center participated in this study. The femoral
   endarterectomy samples were collected from the participants in the
   surgery room and immediately frozen in liquid nitrogen for histological
   examination. Tissue fixation, H&E staining, and immunofluorescence
   staining were performed as mentioned in the previous sections.

Statistical analysis

   Results are expressed as the mean ± SEM. One-way analysis of variance
   (ANOVA) followed by Dunnett's multiple comparison test was used to
   determine the significance of the experimental results. Obtained data
   were analyzed using GraphPad Prism version 5.5.01. (GraphPad Software,
   USA). Statistical significance was set at P < 0.05.

Results

Common DEGs for each GEO dataset

   The DEG analysis data for individual datasets are provided in
   [79]Supplementary Table 2-5. Six genes, common among the individual
   dataset, were identified (Figure [80]1); the three upregulated genes
   were adenosine monophosphate deaminse3 (AMPD3), PCK2, and regulator of
   G protein signaling 1 (RGS1); the three down-regulated genes were four
   and a half LIM domains 1 (FHL1), luminin subunit alpha 5 (LAMA5), and
   KIAA1671 ([81]Figure S1).

Figure 1.

   [82]Figure 1
   [83]Open in a new tab

   Analysis of differentially expressed genes (DEGs) in each dataset. (A)
   Venn diagram of the DEGs identified in each Gene Expression Omnibus
   dataset. The four datasets shared six DEGs, including three upregulated
   and three downregulated genes. (B) Violin plots from each dataset for
   PCK2 and RGS1 validated in wire-injured femoral arteries. Violin plots
   indicating PCK2 and RGS1 expression levels in atheromatous plaques and
   macroscopically intact tissues.

Expression of PCK2 and RGS1 is associated with neointima hyperplasia in
wire-injured vasculature

   To assess whether the six DEGs were associated with neointimal
   hyperplasia, we evaluated the wire injuries in the C57BL/6J mice. At 4
   weeks post-wire injury, the endothelium of the femoral artery was
   denuded and proliferation and migration of VSMCs resulted in extensive
   neointimal hyperplasia. To identify the cell types that expressed the
   DEGs in the blood vessel, we performed co-localization studies of
   α-SMA, a VSMC marker. Additionally, the expression of DEGs in
   non-injured media and injured media was compared by immunofluorescence
   (Figure [84]2A and [85]Figure S2). Of the three up-regulated DEGs, the
   expression levels of PCK2 and RGS1 were higher in the injured media
   than that in the non-injured media. Contrarily, expression levels of
   AMPD3 were lower in the injured media than in the non-injured media.
   However, all the downregulated DEGs showed higher expression in the
   injured media than in the non-injured media. Both PCK2 and RGS1
   colocalized with α-SMA in neointimal hyperplasia (Figure [86]2A).

Figure 2.

   [87]Figure 2
   [88]Open in a new tab

   Involvement of PCK2 and RGS1 in vasculature injury. (A) PCK2 and RGS1
   in the non-injury and injury media and neointima were stained with
   anti-PCK2 antibody and anti-RGS1 antibody, respectively. The protein
   levels of non-injured and injured media were compared. Vascular smooth
   muscle cells (VSMCs) were stained with anti-α-SMA antibody. Dashed line
   defines the media and neointima layer. Images are representative of 5-6
   independent experiments. (B) VSMCs were transfected with negative
   control, PCK2, or RGS1 siRNA for 48 h. The protein levels of PCK2 and
   RGS1 were determined by western blotting using β-actin as an internal
   control. Blots are representative of four independent experiments.
   Relative band intensities are expressed as the means ± SEMs. ***P <
   0.001 vs. corresponding value in negative control. (C and D) VSMCs were
   transfected with negative control, PCK2, and RGS1 siRNA for 48 h. These
   cells were treated with PDGF (10 ng/ml) for 24 h, and cell
   proliferation was determined using the MTT assay. Relative cell
   proliferation determined using the MTT and cell counting assays was
   expressed as means ± SEMs of four independent experiments. *P < 0.05
   and ***P < 0.001 vs. corresponding value in PDGF-untreated negative
   control, #P < 0.05 and ###P < 0.001 vs. corresponding value in
   PDGF-treated negative control. (E) Protein expression of cell
   cycle-related antigens, PCNA and Ki67, was measured by western
   blotting. VSMCs were transfected with negative control or PCK2 siRNA
   for 48 h and stimulated with PDGF (10 ng/ml) for 24 h. The protein
   levels of PCK2, PCNA, and Ki67 were determined using β-actin as an
   internal control. Blots are representative of three independent
   experiments. Relative intensities were expressed as the means ± SEMs.
   **P < 0.01 and ***P < 0.001 vs. corresponding value in PDGF-untreated
   negative control, ##P < 0.01 and ###P < 0.001 vs. corresponding value
   in PDGF-treated negative control. (F) VSMCs were pre-treated with
   PEPCKi (10 µM) for 24 h and subsequently with PDGF (10 ng/ml) for 24 h.
   Cell proliferation was estimated by the MTT assay. Relative cell
   proliferation is expressed as the means ± SEMs of six independent
   experiments. ***P < 0.001 vs. corresponding value in PDGF-untreated
   control, ###P < 0.001 vs. corresponding value in PDGF-treated control

Inhibition of PCK2 attenuates PDGF-induced VSMC proliferation

   Upon investigating the role of PCK2 and RGS1 in VSMC proliferation,
   PCK2 and RGS1 protein expression was attenuated in siRNA-transfected
   human VSMCs (hVSMCs) (siRNA concentration, 200 nM, 48 h; Figure
   [89]2B). MTT assay was performed to investigate PDGF (10 ng/mL for 24
   h) induced VSMC proliferation. Notably, PDGF-induced increase in VSMC
   proliferation was considerably attenuated by PCK2 siRNA, but not by
   RGS1 siRNA in MTT assay (Figure [90]2C) and was evident from cell
   counting (Figure [91]2D). PDGF-induced expression of the cell
   cycle-related antigens, PCNA and Ki67, was also substantially
   attenuated in hVSMCs transfected with PCK2 siRNA (Figure [92]2E). The
   PCK2 inhibitor, PEPCKi (10 μM), attenuated PDGF-induced proliferation
   of hVSMCs (Figure [93]2F).

PCK2-related genes and signaling pathways

   Using DEG analysis, we identified 314 genes potentially associated with
   PCK2, of which, 167 were upregulated and 147 were downregulated
   ([94]Table S6 and S7). We identified five pathways that were
   considerably enriched with DEGs in the KEGG pathway enrichment analysis
   ([95]Table S8). These pathways included the PI3K-Akt signaling pathway,
   FoxO signaling pathway, calcium signaling pathway, IL-17 signaling
   pathway, and microRNAs in cancer. The detailed characteristics of these
   signaling pathways are listed in [96]Table S8. PCK2 was involved in
   both PI3K-Akt and FoxO signaling pathways.

The Akt-FoxO-PCK2 signaling pathway in PDGF-induced VSMC proliferation

   To investigate the PI3K-Akt pathway of PDGF-induced PCK2 expression,
   hVSMCs were pretreated with pharmacological inhibitors of PI3K/Akt
   pathways, including PI3K (LY294002, LY) and Akt (Akt inhibitor IV, AI)
   inhibitors for 1 h and stimulated with platelet-derived growth factor
   (PDGF) (10 ng/ml) for 24 h. The protein and mRNA levels of PCK2 in
   PDGF-treated cells were inhibited by LY and AI (Figure [97]3A).
   Moreover, PDGF-induced PCK2 expression was inhibited in VSMCs
   transfected with siRNA against Akt2 and Akt3, but not in those
   transfected with siRNA against Akt1 (Figure [98]3B‒D). We also
   investigated PDGF-induced FoxO phosphorylation in the PCK2-associated
   signaling pathway. Likewise, VSMCs were pretreated with LY and AI for 1
   h and stimulated with PDGF (10 ng/ml) for 24 h. Phosphorylation of
   FoxO1 and FoxO3 in PDGF-treated cells was inhibited by LY and AI, but
   not by FoxO4 (Figure [99]3E). VSMCs were transfected with siRNA against
   Akt2 and Akt3 for 48 h and stimulated with PDGF for 24 h.
   Phosphorylation of FoxO1 and FoxO3, increased by PDGF, was considerably
   inhibited in Akt2- and Akt3-depleted cells (Figure [100]3F and G). We
   also transfected VSMCs with siRNA against FoxO1 and FoxO3 for 48 h and
   stimulated them with PDGF for 24 h. PCK2 expression, increased by PDGF,
   was substantially inhibited FoxO1- and FoxO3-depleted cells (Figure
   [101]3H and I). These results suggested that PDGF increased PCK2
   expression via PI3K/Akt and FoxO signaling pathways, particularly via
   Akt2, Akt3, FoxO1, and FoxO3. To assess subcellular localization of
   PCK2, we extracted cytosolic and mitochondrial proteins from hVSMCs
   treated with PDGF (Figure [102]4A) and performed immunofluroscence
   analysis (Figure [103]4B). Interestingly, PCK2 expression increased in
   both cytosol and mitochondria in hVSMCs treated with PDGF.

Figure 3.

   [104]Figure 3
   [105]Open in a new tab

   Identification of the role played by PCK2 in PDGF-induced proliferation
   of vascular smooth muscle cells. (A) Vascular smooth muscle cells
   (VSMCs) were pre-treated with LY (10 µM) or AI (10 µM) for 1 h, and the
   protein and mRNA levels of PCK2 were determined by western blotting and
   RT-PCR, respectively. β-actin and GAPDH were used as internal controls.
   Blots are representative of five independent experiments. Relative
   intensities are expressed as the mean ± SEMs. **P < 0.01 vs.
   corresponding value in control, #P < 0.05, ##P < 0.01, and ###P < 0.001
   vs. corresponding value in vehicle. (B-D) VSMCs were transfected with
   negative control, Akt1, Akt2, or Akt3 siRNA for 48 h, and then
   stimulated with PDGF (10 ng/ml) for 24 h. The protein levels of PCK2
   were determined by western blotting using β-actin as an internal
   control. Blots are representative of three independent experiments.
   Relative intensities are expressed as the means ± SEMs. **P < 0.01 vs.
   corresponding value in PDGF-untreated negative control, ###P < 0.001
   vs. corresponding value in PDGF-treated negative control. (E) VSMCs
   were pre-treated with LY (10 µM) or AI (10 µM) for 1 h, and then the
   protein levels of phospho-FoxO1, phospho-FoxO3, and phospho-FoxO4 were
   determined by western blotting. Total-FoxO1, total-FoxO3, and
   total-FoxO4 were used as internal controls. Blots are representative of
   six independent experiments. Relative intensities are expressed as the
   mean ± SEMs. *P < 0.05, **P < 0.01, and ***P < 0.001 vs. corresponding
   value in control, ##P < 0.01 and ###P < 0.001 vs. corresponding value
   in vehicle. (F and G) VSMCs were transfected with negative control,
   Akt2, or Akt3 siRNA for 48 h and then stimulated with PDGF (10 ng/ml)
   for 24 h. The protein levels of phospho-FoxO1 and phospho-FoxO3 were
   determined by western blotting using total-FoxO1 and total-FoxO3 as
   internal controls. Blots are representative of six independent
   experiments. Relative intensities are expressed as the means ± SEMs.
   **P < 0.01 vs. corresponding value in PDGF-untreated negative control,
   ##P < 0.01 and ###P < 0.001 vs. corresponding value in PDGF-treated
   negative control. (H and I) VSMCs were transfected with negative
   control, FoxO1 or FoxO3 siRNA for 48 h, and then stimulated with PDGF
   (10 ng/ml) for 24 h. The protein levels of PCK2 were determined by
   western blotting using β-actin as an internal control. Blots are
   representative of six independent experiments. Relative intensities are
   expressed as the means ± SEMs. **P < 0.01 and ***P < 0.001 vs.
   corresponding value in PDGF-untreated negative control, ##P < 0.01 and
   ###P < 0.001 vs. corresponding value in PDGF-treated negative control.

Figure 4.

   [106]Figure 4
   [107]Open in a new tab

   Intracellular localization of PCK2 on PDGF stimulation. (A) Protein
   level of PCK2 in cytosol and mitochondria of hVSMCs. PCK2 in the
   cytosolic fraction (cyto) was analyzed by western blot analysis.
   β-tubulin was used as an internal control for cytosolic protein. PCK2
   in the mitochondrial fraction (mito) was analyzed by western blot
   analysis. COXIV protein served as an internal control for mitochondrial
   protein. Quantitative analysis of PCK2 protein expression in cytosolic
   and mitochondrial fractions. Experiments shown are representative of
   three independent experiments. (B and C) PCK2 localizes to the cytosol
   and mitochondria in PDGF-stimulated hVSMCs. Intracellular localization
   of PCK2 via immunocytochemistry. Cells were treated with 10 ng/mL PDGF
   for 24 h, and stained. Cytosol was detected with anti-β-tubulin (red)
   and PCK2 (green) antibodies. Mitochondria were visualized using
   MitoTracker (red) and PCK2 (green) staining. A digitally merged image
   of cytosol and mitochondria signals reveals colocalization of PCK2.
   Images are representative of four independent experiments.

Attenuation of VSMC proliferation in wire-injured vasculature of Pck2
knock-out mouse

   We investigated the role of PCK2 in the formation of neointimal
   hyperplasia in vivo. Blood flow in wire-injured femoral arteries was
   higher in Pck2 knockout mice than that in WT mice 3 weeks post-injury
   (Figure [108]5A). Doppler images of blood flow over time (before
   wire-injury, and 7, 14, and 21 days after wire-injury) are shown. H&E
   staining and immunofluorescence analysis revealed that Pck2 knockout
   remarkably reduced neointimal hyperplasia (Figure [109]5B).

Figure 5.

   [110]Figure 5
   [111]Open in a new tab

   Identification of pivotal role that PCK2 plays in vascular restenosis
   and atherosclerosis. (A) Time courses of changes in blood flow in
   wire-injured femoral arteries of WT and Pck2 KO mice. Blood flow in the
   femoral arteries of WT and Pck2 KO mice at 0 (before wire-injury), 1,
   2, and 3 weeks after injury was monitored using a laser Doppler
   perfusion imaging (LDPI). Arrows indicate blood flow in injured femoral
   arteries. Photographs are representative of five independent
   experiments. (B) Cross sections of mouse femoral arteries were prepared
   at four weeks after wire injury and stained with H&E. Vascular smooth
   muscle cells were stained with anti-α-SMA antibody. Dashed line defines
   media and neointima layer. Images are representative of six independent
   experiments. (C) Human femoral atheroma retrieved when performing
   endarterectomy of the common femoral artery were stained with
   anti-α-SMA and PCK2 antibodies. Images are representative of five
   independent experiments.

PCK2 expression in VSMCs of human femoral atheroma

   To examine the expression of PCK2 in human femoral atheroma, we
   collected five femoral atheroma samples during endarterectomy of the
   common femoral artery. The proliferation of VSMCs and colocalization of
   PCK2 with α-SMA in all femoral atheroma samples was observed (Figure
   [112]5C).

Discussion

   We identified six DEGs from the four gene expression datasets. Of the
   six DEGs, only PCK2 and RGS1 displayed identical expression patterns in
   wire-injured femoral arteries of mice and the gene expression datasets.
   Moreover, PDGF-induced VSMC proliferation was attenuated in hVSMCs
   transfected with PCK2 siRNA, but not in cells transfected with RGS1
   siRNA. Notably, RNA sequencing of hVSMCs treated with PCK2 siRNA
   revealed that PCK2 was associated with both the PI3K-Akt and FoxO
   signaling pathways. We also demonstrated that Akt2, Akt3, FoxO1, and
   FoxO3 were associated with PDGF-induced PCK2 expression in hVSMCs. In
   addition, PCK2 had a direct effect on VSMC proliferation when it was
   inhibited with an inhibitor. Remarkably, wire injuries in Pck2-KO mice
   had decreased neointimal hyperplasia and increased blood flow to the
   injured extremity as observed by Doppler imaging. Lastly, we also
   demonstrated that PCK2 was expressed in proliferating VSMCs in human
   femoral atherosclerotic plaques.

   High-throughput technologies have been indispensable tools for
   investigating candidate biomarkers and biological pathways [113]^23.
   Large-scale investigations of transcriptional changes related with
   biological situations of interest have made it possible to identify
   differentially expressed genes associated with diseases [114]^24^,
   [115]^25. However, differences in laboratory protocols, analytical
   methods, and small sample sizes have led to low reproducibility and
   consistency between studies. Independent studies have also attributed
   this discrepancy to the heterogeneity present in many disease subtypes
   [116]^26. Therefore, to overcome these limitations, we identified
   common DEGs from different datasets. Of these, two datasets included
   intact and atherosclerotic carotid arteries (or plaques), whereas
   others included early or stable atherosclerotic plaques and advanced or
   unstable atherosclerotic plaques in the carotid arteries. Using this
   strategy, we narrowed down candidate genes involved in disease
   exacerbation (i.e., advanced carotid plaque lesions and plaque rupture)
   by selecting genes that were differentially expressed between normal
   and disease samples. Thereafter, we investigated which candidate genes
   were involved in VSMC proliferation that had not been studied by the
   authors of the datasets that were analyzed in this study.

   We demonstrated that the Akt-FoxO-PCK2 pathway was associated with
   PDGF-induced VSMC proliferation, particularly via Akt2, Akt3, FoxO1,
   and FoxO3. The PI3K/Akt/FoxO signaling pathway is crucial for multiple
   cellular functions, including cell proliferation and survival [117]^27.
   Reportedly, knockdown of Akt1 on an apolipoprotein E knockout
   background (ApoE^-/- Akt1^-/-) increased the levels of inflammatory
   mediators and reduced eNOS phosphorylation, leading to severe
   atherosclerosis and occlusive coronary artery [118]^28. Deletion of
   Akt1 also reduced VSMC proliferation and migration, thereby inducing
   plaque vulnerability [119]^29. Rensing et al. discovered that ApoE^-/-
   Akt2^-/- double knockout mice exhibited smaller but more complicated
   carotid atherosclerotic plaques [120]^30. Additionally, they
   demonstrated that Akt2 knockout impaired collagen synthesis, VSMC
   proliferation, and migration and altered the expression of
   metalloproteinases and tissue inhibitors of metalloproteinases.
   Although it is mainly expressed in the brain, Akt3 also has a
   protective function in atherosclerosis. Ding et al. demonstrated that
   genetic ablation of Akt3 in ApoE^-/- mice led to a two-fold increase in
   atherosclerotic lesions and that Akt3 suppressed foam cell formation by
   specifically inhibiting the accumulation of macrophage cholesteryl
   esters [121]^31.

   FoxOs induce the expression of pro-apoptotic members of the Bcl2-family
   and inhibit cell growth and apoptosis. Therefore, FoxOs have been
   studied as a key class of tumor suppressors in a range of malignancies
   [122]^32. Notably, Akt directly inhibits FoxOs by phosphorylation and
   regulates cell survival, growth, and proliferation. Previous studies
   have shown that Akt1 plays a protective role against atherosclerosis in
   VSMCs and that FoxO3a activity is elevated in VSMCs atherosclerotic
   plaques in human [123]^29^, [124]^33. In vivo experiments with
   transgenic mice expressing Akt1 specifically activated in arterial
   VSMCS demonstrated that Akt1 inhibited apoptosis in VSMCs. Moreover, it
   inhibited negative remodeling after carotid ligation through FoxO3a and
   Apaf1 [125]^34. Yu et al. demonstrated that FoxO3a activation promoted
   the expression of MMP13 and induced apoptosis in VSMCs and breakdown of
   the extracellular matrix [126]^35. Additionally, Deng et al. reported
   that Akt up-regulates a key osteogenic transcription factor,
   runt-related transcription factor 2, and induces VSMC calcification via
   FoxO1/3 [127]^36. In our study, both Akt2 and Akt3 increased the
   phosphorylation of FoxO1 and FoxO3. In addition, we demonstrated that
   knockdown of FoxO1 and FoxO3 in VSMCs decreased the expression of PCK2.
   Therefore, our results are consistent with those of previous studies,
   i.e., the proapoptotic function of FoxO1/3 in VSMCs is inhibited by
   Akt2/3. Moreover, we elucidated that PCK2 is a downstream component of
   the Akt-FoxO pathway that regulates VSMC proliferation. Given that Akt1
   is expressed predominantly in VSMCs [128]^27 and the roles of Akt2 and
   Akt3 in VSMC proliferation have not yet been thoroughly studied
   [129]^37, this study is distinctive in that it investigates the effects
   of Akt2/3-FoxO1/3-PCK2 pathway on VSMC proliferation.

   Phosphoenolpyruvate carboxykinase (PEPCK) has a key role in
   gluconeogenesis. Although the role of PEPCK-C (cytoplasmic form; PCK1)
   has been thoroughly investigated in diabetes and obesity [130]^38, that
   of its counterpart isozyme, PEPCK-M (mitochondrial form, PCK2), remains
   unexplored. Recent reports have revealed that a potential role of PCK2
   is to promote tumor growth and cell proliferation. These studies
   indicated that glucose restriction enhanced phosphoenolpyruvate
   synthesis from glutamine via PCK2 [131]^39^, [132]^40. Moreover, PCK2
   expression was necessary for the maintenance of tumor cell
   proliferation in vitro under glucose-limiting circumstances and for
   tumor growth in vivo.[133]^41. Ma et al. reported that RNA-binding
   protein Lin28a regulates cardiomyocyte hypertrophic growth by
   increasing PCK2 transcripts and facilitating metabolic repatterning in
   response to cardiac stress [134]^42. VSMCs are dependent on glycolysis
   rather than mitochondrial glucose oxidation [135]^43. When switched
   from the contractile phenotype to the synthetic phenotype of VSMCs in
   atherosclerotic plaques, their dependency on glycolysis increases with
   a decrease in glucose oxidation [136]^44. It was reported that the
   expression of hexokinase 2, an enzyme that catalyzes phosphorylation of
   glucose to glucose-6-phoshate, led to a three-fold increase in the
   human carotid atherosclerotic plaque cap, which may contribute to
   increased glycolytic activity and decreased oxidative phosphorylation
   in VSMCs [137]^45. To date, little attention has been paid to PCK2's
   role in metabolic adaptation on VSMC proliferation.

   Wire injury in an animal model reproduces atherosclerotic conditions,
   such as neointimal formation post-stent implantation, balloon
   angioplasty, or endarterectomy, owing to limitations in conducting
   clinical studies [138]^46. In our in vivo study, intimal hyperplasia in
   wire-injured femoral artery was substantially attenuated in
   Pck2-deficient mice compared with that in WT control mice. Doppler
   imaging indicated that a wire-injured Pck2 KO mouse exhibited improved
   arterial perfusion in the injured extremity compared with wire-injured
   wild-type mice. Moreover, we demonstrated that PCK2 levels were
   measurable in human atherosclerotic plaques of the femoral arteries.
   Interestingly, VSMCs in human atherosclerotic plaques had the most
   expression of PCK2, suggesting that PCK2 can be utilized as a
   therapeutic target for preventing restenosis after interventional
   treatment of peripheral atherosclerotic lesions. This study has some
   limitations. First, in silico analysis, the DEGs were derived from
   datasets of atherosclerotic lesions. We validated their expression in
   wire-injured femoral arteries in mice. It is possible that the DEGs
   other than PCK2 and RGS1 were not validated because changes following
   wire-injury are acute phase and atherosclerotic lesions are chronic.
   Second, we did not elaborate on how PCK2 affects VSMC proliferation,
   specifically as a metabolic adaptation. It should be the goal of our
   future studies.

Conclusion

   Although the use of stents has improved the patency rates in
   femoropopliteal disease, nearly 30-40% of patients experience in-stent
   restenosis within 2 years of implantation [139]^47. To reduce
   neointimal hyperplasia, local delivery of drugs via drug-coated
   balloons or drug-coated stents has been a widely used approach. In the
   clinical scenario, drug-eluting balloon angioplasty presents a superior
   outcome to uncoated balloon angioplasty for the treatment of in-stent
   restenosis of the femoropopliteal arteries [140]^48. However, it is
   necessary to develop new therapeutic targets to prevent restenosis and
   to ensure an optimal clinical outcome. In this study, we present
   evidence that PCK2 is primarily expressed by VSMCs and plays a
   protective role in atherosclerotic injury. Pharmacodynamic techniques
   are evolving to activate specific downstream signaling pathways to
   optimize treatment efficacy and reduce side effects, such as
   immune-mediated disorders [141]^49. Therefore, our results suggest that
   targeting PCK2, a downstream component of the Akt-FoxO-PCK2 pathway, is
   a potential valuable strategy for the treatment of atherosclerotic
   lesions associated with vascular injury.

Supplementary Material

   Supplementary figures.
   [142]Click here for additional data file.^ (836.2KB, pdf)

   Supplementary tables.
   [143]Click here for additional data file.^ (23.3MB, xlsx)

Acknowledgments