Disruption of phactr-1 pathway triggers pro-inflammatory and pro- atherogenic factors: New insights in atherosclerosis development

Abstract

Significant interest has recently emerged for phosphatase and actin regulatory protein (PHACTR1) gene in heart diseases prognosis. However, the functional role of phactr-1 protein remains elusive in heart related-diseases such as atherosclerosis, coronary artery calcification, ischaemic stroke, coronary artery stenosis and early-onset myocardial infarction. Phactr-1 is directly regulated by vascular endothelial growth factor A165 (VEGF-A165) through VEGF receptor 1 (VEGR-1) and Neuropilin-1 (NRP-1). Using an antagonist peptide approach to inhibit the interaction of VEGF-A165 to NRP-1 and VEGF-R1, we high- lighted the importance of both cysteine residues located at the end of VEGF-A165 exon-7 and at the exon- 8 to generate functional peptides, which decreased Phactr-1 expression. Here, we report original data showing Phactr-1 down-expression induces the expression of Matrix Metalloproteinase (MMP) regula- tors such as Tissue inhibitor of metalloproteinase (TIMP-1/-2) and Reversion-inducing-cysteine-rich protein with kazal motifs (RECK). Furthermore, focal adhesion kinases (FAK/PYK2/PAXILLIN) and meta- bolic stress (AMPK/CREB/eNOS) pathways were inhibited in endothelial cells. Moreover, the decrease of phactr-1 expression induced several factors implicated in atherosclerotic events such as oxidized low- density lipoprotein receptors (CD36, Clusterin, Cadherin-13), pro-inflammatory proteins including Thrombin, Thrombin receptor 1 (PAR-1), A Disintegrin And Metalloprotease domain-9/-17 (ADAM-9/-17), Trombospondin-2 and Galectin-3. Besides, Phactr-1 down-expression also induces emerging athero- sclerosis biomarkers such as semicarbazide-sensitive amine oxidase (SSAO) and TGF-beta-inducible gene h3 (bIG-H3). In this report, we show for the first time the direct evidence of the phactr-1 biological function in the regulation of pro-atherosclerotic molecules. This intriguing result strengthened heart diseases PHACTR-1 single-nucleotide polymorphisms (SNP) correlation. Taken together, our result highlighted the pivotal role of phactr-1 protein in the pathogenesis of atherosclerosis.

1. Introduction

Phactr-1 is a newest protein recently discovered in mammalian cells. Phactr-1, a highly conserved protein, belongs to phosphatase and actin regulatory proteins family, which comprises four mem- bers (Phactr-1 to -4) able to bind phosphatases 1 (PP1) and actin with c-terminal RPEL domains. Phactr-1 has been initially identi- fied in rat tissues such as central nervous system, lung, kidney, testis and heart [1]. Interestingly, single-nucleotide polymorphisms (SNP) at the locus Chr6p24.1 corresponding to PHACTR1 was asso- ciated with myocardial infarction [2e4], atherosclerosis [5], coro- nary artery calcification [2,4,6], ischaemic stroke [7], coronary artery stenosis [8], coronary heart disease in type 2 diabetes [8] and early-onset myocardial infarction [9]. The physiological role of phactr-1 remained obscure, however recent advance in this field showed that phactr-1 is involved in angiogenesis process including tube formation and lamellipodial dynamics. In these biological processes, phactr-1 expression is regulated and controlled by VEGF- A165 through its interaction with neuropilin-1 (NRP-1) and VEGF- R1 [10].

NRP-1 was initially identified as neuronal receptors of specific secreted members of semaphorin III family involved in guidance and axonal repulsion [11e13]. Furthermore NRP-1 is preferentially expressed in arteries [14,15]. The disruption or overexpression of NRP-1 is lethal in mice due to vascular abnormalities or excess of vessel formation respectively [16,17]. However, NRP-1 expression persists in adult particularly in heart [18] and placenta [19], but also in smooth muscle cells [20], epithelial cells [21], immune cells [22,23] and endothelial cells [24]. NRP-1 plays also a pivotal role in vascular system as receptors of some members of VEGF (vascular endothelial growth factor) family, to modulate the VEGF/VEGF- Remediated signalling. In addition, to modulate its biological functions, VEGF binds different receptors such as the tyrosine ki- nase receptors VEGF-R1/-R2/-R3, with various affinities [25]. VEGF- R1 has a pivotal role in the modulation of endothelial precursors and mature monocytes/macrophages migration. VEGF-R2 is the main signal transducing VEGF-A in endothelial cells [25,26] and VEGF-R3 is particularly involved in lymphatic function [27].

Here, we strengthened the knowledge of Phactr1 protein function. We report that down-expression of phactr-1 enhances Matrix Metalloproteinase (MMP) regulators, which inhibit FAK/PYK2/ PAXILLIN pathways. Moreover, AMPK/CREB/eNOS is also altered. Furthermore, phactr-1 down-expression induced several factors implicated in atherosclerosis events such as LDL-oxidized receptors (CD36, Clusterin, Cadherin-13), pro-inflammatory proteins (Thrombin, PAR-1, Trombospondin-2, GAL-3). Besides, phactr-1 inhibition also induces emerging atherosclerosis biomarkers (SSAO, b-IGH3).

2. Materials and methods

2.1. Chemical peptide synthesis

Peptide synthesis was carried out by solid phase on an A433 synthesizer (Applied Biosystems) following the standard small scale (0.1 mmol) Fmoc chemistry. Rink MBHA resin, Fmoc-Arg(Pbf) or Fmoc-Cys(Trt) preloaded Wang resin, Fmoc-Lys(Fmoc)-OH, Fmoc-Ahx-OH and all the other suitably protected amino acids were purchased from Novabiochem (Germany). Amino acids were used in 10 equivalents (1 mmol), HBTU/HOBt/DIPEA (1 mmol/ 1 mmol/2 mmol) were used as coupling agents, 20% piperidine as Fmoc deprotection agent, NMP as solvent. For the synthesis of branched peptides (p13), 0.05 mmol or 0.025 mmol resin was used to ensure complete coupling of each step, as the dimeric chains were introduced simultaneously on lysine residues. After synthesis,
each peptide was cleaved from the polymer support and freed from side-chain protections in trifluoroacetic acid/water/triisopropylsi- lane (9.5:0.25:0.25 in volumes) or in trifluoroacetic acid/water/ triisopropylsilane/1,2-ethanedithiol (9.4/0.25/0.1/0.25 in volumes) for cysteine-containing peptides for 4 h and then precipitated in cold diethyl ether after evaporation. Crude peptide was collected by centrifugation and purified by semi-preparative reversed-phase HPLC on a Vydac C18 column (10 × 250 mm) with acetonitrile/ water containing 0.1% trifluoroacetic acid. The peptide fractions were collected, lyophilized and analysed by reverse-phase HPLC on a Vydac C18 column (4.6 × 250 mm). The molecular weight of peptides was verified by ion electrospray mass spectrometry. Peptide stock solutions (10 mM) were prepared in an aqueous solution containing DTT (100 mM) in order to avoid cysteine-induced peptide dimerization. These solutions were stored at —20 ◦C.

2.2. Cell culture conditions

Primary human umbilical vein endothelial cells (HUVEC) were cultured at 37 ◦C and 5% CO2 in EGM-2 medium, supplemented with 10% foetal bovine serum and growth factors (all from Lonza, Belgium). HUVEC cells between passages 3 and 4 were used for all experiments. Stock solutions (10 mM) of tested peptides were prepared in DTT (100 mM) and stored at —20 ◦C. The final peptide concentrations in culture were 10 mM.

2.3. Total RNA preparation, RT-PCR and RQ-PCR

HUVEC RNA was extracted with NucleoSpinRNA II kit (Macher- eyeNagel, France) and quantified using Nanodrop (ND-1000 spectrophotometer). 1 mg of each RNA sample was reverse- transcripted into cDNA using iScript cDNA Synthesis Kit (Bio-Rad, USA) following the manufacturer’s instructions. Specific primers were used: PHACTR1, 5′-GAG-GCA-AAG-CAG-AGA-AGA-GC-3′ and 5′-CAT-GAT-GTC-TGACGG-TTG-GA-3′; RPLO, 5′-CAT- TGCeCCCeATG-TGA-AGT-C-3′ and 5′-GCTeCCCeACT-TTG-TCTCCA-GT-3′. PCR amplification was performed in reaction mixture (25 mL) containing 200 mM of each dNTP, 1 mg of cDNA, 1 mM of
primers and 0.625U of GoTaq DNA Polymerase (Promega, France) with 45 s of denaturation at 95 ◦C, 45 s of annealing at 60 ◦C and 1 min of extension at 72 ◦C for 30 cycles. PCR products were separated by 1% agarose gel electrophoresis, stained with ethidium bromide (Sigma, Germany) and analysed using Gel Doc 2000 Sys- tem (Bio-Rad, USA).

PCR amplifications were performed using the Real-Time PCR Systems (Applied Biosystem, USA) and the IQ™ SYBR® Green Supermix kit (Biorad, USA) according to the manufacturer’s pro- tocol. Two microlitres of the first strand cDNA (1:5 diluted) were used for amplification in triplicate in a 10 mL reaction solution contained SYBR® Green and 10 pmol of each primer. The following PCR program was used: 95 ◦C for 3 min, 45 amplification cycles at
95 ◦C for 10 s, 60 ◦C for 30 s. Serial dilutions of each studied tran- script were used to determine the amplification efficiency of each target and housekeeping gene (RPLO).

In the present study, data are obtained by analysis with BioRad MFX Software 2.0 and are presented as the fold-change in target gene expression normalized to the internal control gene. The average threshold cycle (CT) was calculated for both the target gene PHACTR1 and RPLO, and DCT was determined as [the mean of the triplicate CT values for the target gene PHACTR1] — [the mean of the triplicate CT values for RPLO]. DDCT represented the difference between the paired samples, as calculated by the formula DDCT = (DCT of sample — DCT of control). The statistical analysis was carried out with the ANOVA test.

2.4. Inhibition of VEGF-A165 binding to NRP-1 and VEGF-R1

NRP-1 or VEGF-R1 (R&D Systems, France) were coated to the surface of a 96-well plate (COSTAR) by incubation at 0.2 mg/mL in PBS at 4 ◦C overnight. The plate was washed three times in PBS containing 0.5% Tween 20 (buffer A), after which it was treated for 2 h at 37 ◦C with PBS containing 0.5% BSA (buffer B) to block non- specific binding and rinsed with PBS. The plate was then incubated with peptides at 50 mL/well (10—4 M) for 1 h at 37 ◦C followed by a 2 h incubation at 37 ◦C with biotinylated VEGF-A165 ((bt)-VEGF-A165), supplemented with 4 mg/mL heparin (all from R&D Systems, France). (bt)-VEGF-A165 was added at 200 mg/mL for NRP-1 and at 100 mg/mL for VEGF-R1. Then 100 mL of streptavidinehorseradish peroxidase (GE Healthcare, Germany) diluted at 1:8000 in buffer A was added. After 1 h incubation at 37 ◦C under obscurity, the plate was washed five times with buffer A and 100 mL of SuperSignal West Pico Chemiluminescent Substrate (Pierce, USA) was added. Luminescence was quantified with an EnVision 2101 Multilabel Reader (Perkin Elmer, USA).

2.5. Tube formation assay

For tube formation assay, the underlying collagen gel, 70 mL of collagen/media solution (for 1 mL containing of 423 mL type I rat tail collagen (Invitrogen), 44 mL 0.1 N NaOH, 48 mL 10X Na2HCO3 and 0.5 mL EBM2 medium supplemented with 2% FBS) was put in each well of 96-well culture plate. After 30 min at 37 ◦C, 5% CO2, HUVEC cells were seeded (8 × 105 cells) on collagen with EBM2 supplemented with 50 ng/mL of VEGF-A165 in absence (control) or
presence (treatment) of the peptide. After confluent monolayer formation, a second collagen gel layer was added over the apical cell surface. After 24 h, tube formation was observed. Wells were photographed using a SPOT camera attached to a Nikon Eclipse TE2000-S inverted microscope. Resulting images were analysed using Wimasis image analysis platform (Ibidi, France).

2.6. Proteome profiler arrays

Biochemical signalling detection was evaluated by using human proteome profiler array (human common analytes array, non- haematopoietic panel and human phosphokinase arrays) according to the manufacturer’s instructions (R&D systems, France). Briefly, capture and control antibodies were spotted in duplicate on nitrocellulose membranes. Cellular extracts were diluted and incubated overnight with the Human proteome Arrays. The array was washed to remove unbound proteins, followed by incubation with a cocktail of biotinylated detection antibodies. Streptavidin- HRP and chemiluminescent detection reagents were applied, and the signal intensity corresponding to the amount of protein bound was measured at each capture spot.

2.7. siRNA transfection

4.5$104 HUVECs were seeded in 6-well plates 24 h prior to transfection. siRNA-mediated knockdown was performed using HiPerfect transfection reagent (Qiagen, France). Three different siRNAs for each target were tested. Here we report only the efficient siRNA sequences (DSIR algorithm): siphactr-1, CGAAGACGACGA- CAGCUCATT was from Dharmakon (USA). siScramble was used as negative control (Qiagen, France). All siRNAs were used at a 10 nM final concentration. Cells were then analysed for phactr-1 and housekeeping rplo mRNA expression.

2.8. Proliferation assays

The cell proliferation assay is a colourimetric assay based on the enzymatic cleavage of the tetrazolium salt (WST-1, Roche, France) to a water-soluble coloured formazan dye by mitochondrial succinate-tetrazolium reductase in viable cells. The exponentially growing cells were seeded in 96-well plates (HUVEC 3 × 103 cells/well). After 72 h, 10 mL of WST-1 reagent was added in the wells and the cells were incubated at 37 ◦C in a humidified atmosphere of 5% CO2for 1e2 h. Optical density (OD) was measured at 490 nm using a microplate reader Manager 5.2 (Bio-Rad, USA).

2.9. Cell adhesion assay

Hexamethyl pararosaniline chloride is used to measure cell viability. HUVEC were seeded in 96-well plates (3000 cells/well). After 24 h, they were treated with different peptides. After 72 h, the adherent cells were fixed with 50 mL 3% paraformaldehyde for 15 min at room temperature. The medium and the non-adherent cells were removed, the cells were washed extensively with 100 mL PBS (with Ca2+ and Mg2+) (Lonza, Belgium) and then stained
with 50 mL of crystal violet 0.04% for 30 min at room temperature.The cells were then washed gently 3 times with PBS 100 mL and were permeabilized with 100 mL triton 2% for 15 min. The micro- plate was stirred to obtain a homogenous solution and the absor- bance was quantified at 580 nm by Microplate Manager 5.2 (Bio- Rad, USA).

2.10. Statistical analysis

Data were expressed as the arithmetic mean ± SD of at least three different experiments. The statistical significance of results was evaluated by one-way ANOVA using GraphPad Prism®, with probability values *p < 0.05, **p < 0.01, ***p < 0.001, being considered as significant. The postdoc test associated with ANOVA was Dunnett's.

3. Results

3.1. VEGF-A165/VEGF-R1/NRP-1 peptidic inhibitory approach

As we previously reported, a peptidic approach using amino- acid sequences derived from VEGF-A165 (the end of exon-7 and exon-8) is crucial to define specific peptide sequences able to antagonize VEGF-A165 binding to NRP-1, VEGF-R1 or both (Fig. 1A, B) [10]. We observed that small peptide sequences containing cysteine residue (p2, ERTCRC-OH; p4, CDKPRR-OH) acquired the efficient capacity to block both VEGF-A165/NRP-1 (88 ± 12% and 96 ± 3%, respectively) (Fig. 1C) and VEGF-A165/VEGF-R1 (64 ± 20% and 88 ± 1%, respectively) (Fig. 1D) compared to the native VEGF- A165 sequence (p1, ERTCRCDKPRR-OH) (97 ± 2% and 17 ± 4%, respectively). Then, a second series of peptides was synthesized by connecting p2 (ERTCRC-OH) and p4 (CDKPRR-OH) sequences using an aminohexanoïc (Ahx) spacer moiety (p5, ERTCRC-(Ahx)- CDKPRR-OH; p6, ERTCRC-(Ahx)2-CDKPRR-OH; p7, ERTCRC-(Ahx)3-
CDKPRR-OH; p8, ERTCRC-(Ahx)4-CDKPRR-OH) in order to increase their own sensitivity (to attempt concentration less than 10—4 M). However, only an increase of VEGF-A165/VEGF-R1 binding inhibi- tory effect has been observed (43 ± 18%, 46 ± 8%, 33 ± 6% and 42 ± 9%, respectively) compared to native VEGF-A165 sequence (p1, ERTCRCDKPRR-OH) (17 ± 4%). Indeed, in contrast to p2 (ERTCRC-OH) and p4 (CDKPRR-OH), these peptides (p5, p6, p7 and p8) have dramatically lost their capacity to block VEGF-A165/VEGF-R1 inter- action but no significant alteration of VEGF-A165/NRP-1 inhibition has been observed (Fig. 1C, D). In conclusion, no newly synthesized peptides showed more potent efficiency than the native VEGF-A165 sequence (p1, ERTCRCDKPRR-OH). Moreover, the efficient concen- tration of all optimized peptides remained in 10—4 M ranges. Originally, we have pointed out the cysteine's importance [28e30] in the peptide sequence to confer VEGF-A165 inhibitory activity to peptides (p1, ERTCRCDKPRR-OH; p2, ERTCRC-OH; p4, CDKPRR-OH)
compare to lacking cysteine sequence peptide (p3, DKPRR-OH) (Fig. 1C, D) [10]. Here, we extended peptide optimization to clarify the cysteine's importance in peptide sequence. Therefore, we developed 3 peptides with successive cysteineeserine substi- tution (p10, ERTCRSDKPRR-OH; p11, ERTSRCDKPRR-OH; p12, ERTSRSDKPRR-OH) (Fig. 1B). This single sequential cysteineeserine substitution (p10, ERTCRSDKPRR-OH; p11, ERTSRCDKPRR-OH)
conferred to peptides an increased capacity to inhibit VEGF-A165/ VEGF-R1 (80 ± 0% and 87 ± 2%, respectively) compared to native VEGF-A165 sequence (p1, ERTCRCDKPRR-OH) (Fig. 1D) in contrast to the inhibition of VEGF-A165/NRP-1 interaction, which remained unchanged (Fig. 1C). However, a full cysteineeserine substitution (p12, ERTSRSDKPRR-OH) abolished VEGF-A165 inhibitory capacity for both receptors (VEGF-R1 and NRP-1) (7 ± 4% and 38 ± 6%, respectively) (Fig. 1C, D). These results underlined the importance of both cysteine residues located at the end of exon-7 as well as at the exon-7/exon-8 junction site but not exclusively at the exon-8.

3.2. Phactr-1 inhibition using antagonist peptides with serineecysteine substitution

The expected results were the identification of full-antagonist peptides able to switch-off phactr-1 mRNA in HUVEC through the total or partial VEGF-A165/VEGF-R1/NRP-1 binding inhibition since this complex regulated phactr-1 expression [10]. We initially identify p1 (ERTCRCDKPRR-OH), p2 (ERTCRC-OH) and p4 (CDKPRR-OH) as fully inhibitors of phactr-1 expression compared to partial antagonist peptide (p3, DKPRR-OH), which barely inhibits VEGF- A165 binding to both, VEGF-R1 and NRP-1 (Fig. 1E, C, D). The sequential cysteineeserine substitution peptides (p10, ERTCRSDKPRR-OH and p11, ERTSRCDKPRR-OH), which antagonized VEGF-A165 binding to both receptors, are also efficient to inhibit phactr-1 expression compared to fully cysteineeserine substituted peptide that did not (p12, ERTSRSDKPRR-OH) (Fig. 1F). Peptide p12 significantly lose its inhibitory property to NRP-1 and VEGF-R1 (Fig. 1C, D). In contrast, the following newly synthesized peptides (p6, ERTCRC-(Ahx)2-CDKPRR-OH; p7, ERTCRC-(Ahx)3-CDKPRR-OH; p8, ERTCRC-(Ahx)4-CDKPRR-OH) are able to inhibit phactr-1 expression but not p5 (ERTCRC-(Ahx)-CDKPRR-OH). p5 in- efficiency might be reliable to the presence of unique Ahx spacer, which brought low flexibility compared to others (Fig. 1G). Finally, we studied a specific VEGF-A165/VEGF-R1 inhibitory peptide (p9, GNQWFI-OH), which also remains non-efficient to reduced Phactr- 1 expression (Fig. 1F).

Taken together, down-expression of phactr-1 using VEGF-A165 binding antagonist peptides is dependent of VEGF-A165/NRP-1 binding inhibition in contrast to our previous VEGF receptors siRNA approach, which showed the implication of both receptors [10]. These results mean that down-modulation of phactr-1 expression is not solely due to the VEGF-A165 binding to NRP-1 and/or VEGF- R1. Nevertheless, the potential implication of NRP-1/VEGF-R1 forming complex remains to be investigated.

3.3. Generation of dimeric peptide from monomeric native sequence and its characterisation

As the native sequence (p1, ERTCRCDKPRR-OH) seemed to be the most efficient to mimic physiological VEGF-A165 binding inhi- bition as well as phactr-1 down-expression, we synthesized the dimeric peptide (p13, (ERTCRCDKPRR-Ahx)2-Lys-NH2) (Fig. 2A). As expected, p13 inhibited the VEGF-A165 binding to NRP-1 with the same p1 monomeric peptide efficiency (98 ± 2% vs 97 ± 2%) (Fig. 2B). However, dimeric peptide (p13) acquired the capacity to block VEGF-A165/VEGF-R1 binding (88 ± 1%) (Fig. 2C). Thus, dimerization of native VEGF-A165 sequence seemed to confer to this peptide a binding antagonist property to VEGF-R1 compared to monomeric p1 VEGF-A165 sequence (17 ± 4%) (p ≤ 0.001). A barely detectable effect in HUVEC proliferation (p ≤ 0.05) but none on HUVEC adhesion has been observed (Fig. 2D, E). Surprisingly, dimeric peptide is also not efficient to maintain phactr-1 expression inhibition (Fig. 2F). Once again, phactr-1 down-expression escapes to peptide optimization when this main peptide sequence became VEGF-A165/VEGF-R1 binding inhibitor. However, to better under- stand mechanism of native peptide effect on phactr-1 expression, we tested both peptides in tube formation assays compared to phactr-1 siRNA approach.

3.4. Phactr-1 down-expression interferes with tube morphology

Monomeric and dimeric peptides showed functional similarity since both are able to block tube formation compared to the un- treated HUVEC (Fig. 3A). Tube structure and morphology study did not point up significant difference between them. In details, these peptides (p1, ERTCRCDKPRR-OH and p13, (ERTCRCDKPRR-Ahx)2-Lys-NH2) were capable to reduce tube numbers, which included cell-covered area, number of branching points and number of loops (p ≤ 0.001) (Fig. 3B). More importantly, inhibition of tube formation is associated with a reduction of total tube length in both peptides-
treated HUVEC. This extended study of phactr-1 down-expression showed that inhibition of tube formation might be done by the blocking of phactr-1 but is not restrictive to it own expression. In- hibition of VEGF-A165 binding to both, NRP-1 and VEGF-R1, is also sufficient to block tube formation that highlighted the alternative blocking pathway.

To reinforce this putative major role of phactr-1, we next used validated phactr-1 siRNA to study its impact in tube formation [10] (data not shown). In accordance with the peptide study described below, the suppression of phactr-1 in HUVEC impaired tube for- mation (Fig. 3C). This is associated with a decrease of tube numbers, which encompassed cell-covered area, of branching point numbers and of loop numbers (p ≤ 0.001) (Fig. 3D). Taken together, phactr-1 down-expression seemed to be a fatal way for tube formation and integrity.

3.5. Phactr-1 suppression increases MMP regulators which inhibit focal adhesion protein phosphorylation

We previously reported that phactr-1 down-expression induced actin polymerization/depolymerization alteration, which mediated lamellipodia instability in endothelial cells [10]. To decipher the molecular pathways and molecules involved in phactr-1 mecha- nism of action, we focused our study on different human profiler proteome arrays (human common analytes array, non-haemato- poietic array and human phosphoekinase arrays) (Supplemental Fig. 1). In phactr-1 down-expressing cells, we observed an in- crease expression of both TIMP metalloproteinase inhibitors TIMP-
1 and TIMP-2 (p ≤ 0.001, respectively) but not of TIMP-4 (Fig. 4A).

As TIMP-2 regulates human capillary tube stabilization [31] and TIMP-1 inhibits microvascular endothelial cell migration [32], we assessed whether focal adhesion proteins such as FAK (Focal Adhesion Kinase), PYK2 and Paxillin might be affected [33]. In these cells, 397Y FAK, 118Y Paxillin and 402Y PYK2 phosphorylations were significantly inhibited (p ≤ 0.001, respectively) (Fig. 4B). Taken together, 397Y FAK and 118Y Paxillin dephosphorylation might be explained by both induced TIMP-1/-2 signalling in phactr-1 siRNA treated cells [32,34]. The inhibition of 118Y Paxillin phosphorylation is known to induce other Matrix Metalloproteinase (MMP) regulators like the Reversion-inducing-cysteine-rich protein with kazal motifs (RECK) [34,35] as observed here (p ≤ 0.01) (Fig. 4C). On the whole, the induction of RECK and TIMP-2 which jointly act as a negative regulator for MMP-9/MMP-2/MMP-14 [36], and Membrane-type MMPs (MT-MMPs), respectively, may lead to the alteration of cellecell and cellematrix interactions. Moreover, dy- namics of actin assembly were corrupted by the significant decreased expression of b-catenin, which was probably degraded by the proteasome since GSK-3a/b phosphorylation was also inhibited (p ≤ 0.001 and p ≤ 0.01, respectively) (Fig. 4D).

3.6. Phactr-1 down-expression induces atherosclerosis-involved factors

Since several genetic studies underlined the PHACTR1 SNP involvement in Atherosclerosis disease, we focused our attention on the regulation of pathologic related factors. This axis is partic- ularly relevant due to the low phactr-1 expression associated to PHACTR1 SNP identification [37]. We have previously shown that phactr-1 inhibition decreased PP1 activity [10], which is known to decrease SERCA/Ca2+ intake dependent process [38e40]. In this way, the lack of PP1 activity may induce Ca2+ deficiency in cells, which may explain the inhibition of 174T AMPKa1 (p ≤ 0.001) phosphorylation but not for phosphorylated 172T AMPKa2 as observed in phactr-1 down-modulated endothelial cells (Fig. 5A). Consequently, 174T AMPKa1 dephosphorylation induced the phos- phorylation inhibition of both 1177S eNOS and 133S CREB (p ≤ 0.001 and p ≤ 0.05, respectively) (Fig. 5B). In summary, phactr-1 down-
expression alters both AMPK-mediated phosphorylation of eNOS and CREB that could lead to the decrease of NO production, glu- cogenesis and fatty acid oxidation, respectively. All together, these results provided a link between phactr-1 expression, metabolic stresses and cardiac dysfunctions. Collectively, AMPK and eNOS co- regulation disturbs endothelial homoeostasis and athero- protective phenotype of endothelial cells [41]. Although thrombin was increased in this context, the restoration of 174T AMPKa1 phosphorylation did not occur (p ≤ 0.001) (Fig. 5C). Nevertheless, its binding to the induced PAR-1 (Thrombin receptor 1) may induce pro-inflammatory microenvironment as it was described in atherosclerosis disease [42] (Fig. 5C). However, no significant amount of IL-8 has been induced in this model (Fig. 5C). Further- more, other pro-inflammatory factors associated to atherosclerosis such as Galectin-3 (GAL-3) [43], Trombospondin-2 (TSP-2) [44],

ADAM-9 and ADAM-17 [45,46] were significantly increased (p ≤ 0.001, p ≤ 0.001, p ≤ 0.001 and p ≤ 0.05, respectively) (Fig. 5D). Likewise, phactr-1 down-expression induced the early atheroscle- rosis marker semicarbazide-sensitive amine oxidase (SSAO) [47] (p ≤ 0.001) (Fig. 5E), which is implicated in both inflammation and LDL-oxidation [47,48] and b-IGH3 [46] (p ≤ 0.001) (Fig. 5E). In addition, LDL oxidized receptors (CD36, Clusterin, Cadherin-13) were significantly increased (10-fold, 2-fold and 2.5-fold, respec- tively) on endothelial cells (p ≤ 0.001) (Fig. 5F). In summary, phactr-1 down-expression seemed to induce a several part of components implicated in atherosclerosis process. These different-
induced factors were closely modulated and are implicated in interconnected regulating cascade that emphasize the pivotal role of phactr-1.

4. Discussion

Human PHACTR1 gene is located on chromosome 6 and encodes a 484-amino acid protein. Multiple transcript variants encoding different isoforms have been found for this gene. Phactr-1 modu- lates PP1 and actin by a direct interaction [1]. Recently, we high- lighted the inducible expression of human endogenous PHACTR1 gene in VEGF-A165-stimulated HUVEC by NRP-1/VEGF-R1 complex signalling. Gene silencing experiments demonstrated that phactr-1 expression in HUVEC abrogated in vitro VEGF-A165-induced tube formation by the induction of apoptosis. However, before apoptosis induction, phactr-1-mediated signalling remained elusive. For this reason, we down-modulated phactr-1 expression using siRNA or peptide approaches in HUVEC to investigate its functional role and its intracellular signalling.

Peptide design to generate more efficient tools to block VEGF- A165 binding to NRP-1/VEGF-R1 with lower concentration remained ineffective. The modification of VEGF-A165 derived pep- tides delineates structural features implicated in both NRP-1 and VEGF-R1 binding and functional activity. Moreover, modifications have been shown to improve peptide activity, including the enhancement of proteolytic stability and/or their potency. The design and synthesis of peptides able to modulate VEGF-A165 binding to NRP-1/VEGF-R1 interaction in an exquisitely selective way will be major to easily study Phactr-1 regulation in contrast to siRNA strategy. Thus, we have actively working on this field. Although no significant progress has been obtained, these results clearly increment the knowledge of the functional VEGF sequence structure.

Successive cysteineeserine substitution conferred to peptides an increase capacity to inhibit VEGF-A165/VEGF-R1, however, a fully cysteineeserine substitution abolished its VEGF-A165 inhibitory ca- pacity for both receptors (VEGF-R1 and NRP-1). Taken together, both cysteines are important to confer receptors binding specificity and not only the cysteine located at the exon-7-exon-8 junction. We surprisingly emphasize, that both cysteine residues located at the end of exon-7 and exon-8 are both majors to induce phactr-1 down- expression. To strengthen the literature that point up the importance of cysteine located at the exon-8 to inhibit VEGF-A165 binding, to block tube formation and phactr-1 expression [10], we showed here that at least one cysteine in the peptide sequence is necessary to obtain functional phactr-1 down-modulator peptides. This result brings a newest perspective for antagonist peptides or small inhib- itory molecules study and design in this field. In this report, we highlighted that tube formation may be also destructed indepen- dently of phactr-1 down-expression through the inhibition of VEGF- A165 binding to both NRP-1/VEGF-R1 using peptide approach. Both functional phactr-1 down-modulator peptides and phactr-1 siRNA are able to disturb tube formation in a same manner.

To date, few results have been generated to elucidate the physiological roles of phactr-1 in angiogenesis. Thus, we focused our study on phactr-1 expression associated to tube formation inhibition. Molecular profiles have been investigated using prote- ome arrays to unravel phactr-1 signalling. Interestingly, phactr-1- down-expression in endothelial cells induces MMP regulators that lead to disrupt tube formation. Moreover, MMP regulators induce inhibition of FAK/PYK2/PAXILLIN phosphorylation that allowed RECK-dependent inhibition of MT1-MMP/MMP-9. Sur- prisingly, phactr-1 down-expression induced a large quantity of pro-atherogenic protein components such as pro-inflammatory factors, LDL-oxidized receptors, SSAO and bIGH3. Furthermore, PP1 inactivation mediates alteration of AMPK/CREB/eNOS path- ways alteration probably via SERCA regulation (Fig. 6).

Surprisingly, phactr-1 function study in tube formation using HUVEC model, brought out the induction of proteins implicated in atherosclerosis. Taken together, this study brings new perspectives for the atherosclerosis knowledge but needs further development in arterial models.Furthermore, several studies highlighted the presence of PHACTR1 SNP in correlation with heart diseases. Even if SNP did not systematically alter protein expression, it may change conforma- tional structure that conduct to biological function alteration (including low protein partner affinity). Thus, extended study might be major in these heart diseases to correlate, PHACTR1 SNP and phactr-1 functional protein synthesis, with the prognostic and disease evolution. Moreover, recent investigation in myocardial infarction revealed a correlation between PHACTR1 SNP (rs 9349379) and phactr-1 protein down-expression as a risk factor [37]. Thus, mechanism of action of the phactr-1 down-expression reported here may open new perspective of research in the field of heart diseases and more particularly in atherosclerosis disease.

5. Conclusion

Our present study shed light for the first time that VEGF-A165 peptidic sequences coded by the end of exon-7 and exon-8 binding specificity to NRP-1 is dependent of amino acid sequences and more particularly of both cysteine residues contained in this sequence. Moreover, phactr-1 down-expression is not limited to a peptide sequence able to specifically inhibit VEGF-A165 binding to NRP-1 but is at least restrictive to a peptide sequence that inhibits VEGF-A165 binding to NRP-1. Thus, phactr-1 down-expression is dependent of the VEGF-A165/NRP-1 binding inhibition but not to VEGF-R1. To date, phactr-1 signalling pathway modulation inducing cytoskeleton rearrangement remained unknown, thus using siRNA approach, we showed here that down-expression of phactr-1 negatively regulate FAK/PYK2/PAXILLIN and AMPK/CREB/ eNOS pathways and induces both MMP regulators (TIMP-1/-2, RECK) expression and several factors implicated TD-139 in atherosclerosis.