Biogenesis of Pro-senescent Microparticles by Endothelial Colony Forming Cells from Premature Neonates is driven by SIRT1-Dependent Epigenetic Regulation of MKK6

Senescent cells may exert detrimental effect on microenvironment through the secretion of soluble factors and the release of extracellular vesicles, such as microparticles, key actors in ageing and cardiovascular diseases. We previously reported that sirtuin-1 (SIRT1) deficiency drives accelerated senescence and dysfunction of endothelial colony-forming cells (ECFC) in PT neonates. Because preterm birth (PT) increases the risk for cardiovascular diseases during neonatal period as well as at adulthood, we hypothesized that SIRT1 deficiency could control the biogenesis of microparticles as part of a senescence–associated secretory phenotype (SASP) of PT-ECFC and investigated the related molecular mechanisms. Compared to control ECFC, PT-ECFC displayed a SASP associated with increased release of endothelial microparticles (EMP), mediating a paracrine induction of senescence in naïve endothelial cells. SIRT1 level inversely correlated with EMP release and drives PT-ECFC vesiculation. Global transcriptomic analysis revealed changes in stress response pathways, specifically the MAPK pathway. We delineate a new epigenetic mechanism by which SIRT1 deficiency regulates MKK6/p38MAPK/Hsp27 pathway to promote EMP biogenesis in senescent ECFC. These findings deepen our understanding of the role of ECFC senescence in the disruption of endothelial homeostasis and provide potential new targets towards the control of cardiovascular risk in individuals born preterm.

Eighteen term (control, gestational age (GA) > 37 weeks, appropriate weight) and twenty-33 nine preterm neonates (GA 24 to 35 weeks with appropriate or small weight for GA), were 34 included. Exclusion criteria were congenital viral infections, major congenital heart or 35 structural brain malformations, genetic abnormalities and metabolic diseases. This research 36 was approved by a local ethic committee Assistance Publique Hôpitaux de Marseille and the 37 study was performed conform the declaration of Helsinki. All the parents have provided 38 written informed consent for the use of cord blood. The patient characteristics are shown in 39 Table S1. 40

Isolation of endothelial colony-forming cells. 41
ECFC were isolated and expanded from mononuclear cell fraction (MNC) obtained from the 42 cord blood of term (CT) and preterm (PT) neonates, cultured and characterized as previously 43 described 1 . ECFC were used between the third and fourth passage. 44

Antibody arrays 45
Conditioned media for antibody array analysis were prepared by washing cells with PBS and 46 incubating them in basal medium for 48h. The conditioned media were collected in a 47 centrifuge tube, and the cells remaining on the dish were counted to normalize conditioned 48 media volumes by cell number. The conditioned media were clarified by brief centrifugation, 49 0.2µm filtered, diluted with a serum-free medium to a concentration equivalent to 1.35 x 10 5 50 cells per 1.3ml, and applied to the antibody arrays (Raybiotech; AAH-CYT-1) as described 51 previously by Freund et al 2 and as recommended by the supplier. The signals were detected 52 using a G-BOX Imaging System (GeneSys) and were analyzed using specific software 53 (GeneTools, Syngene). Signals were averaged and displayed as described in the figure  54 legend. 55

ELISA 56
Conditioned media were prepared by incubating cells for 48h as described above. The CM 57 were analyzed using the Human IL-6 ELISA Kit II and reagents, following the procedures 58 described by the manufacturer (BD OptEIATM; 550799). 59 Counting and measuring particles. 60 i.
Tunable resistive pulse sensing. The concentration and size distribution of particles 61 was analyzed with TRPS (qNano, Izon Science Ltd, Christchurch, New Zealand), a 62 relatively new technology that allows the detection of particles passing through a 63 nanopore by way of single-electrophoresis 3 . The technology is based on the Coulter 64 principle at the nano scale and operates by detecting transient changes in the ionic 65 current generated by the transport of the target particles through a size-tunable 66 nanopore in a polyurethane membrane. Samples were diluted to half in PBS buffer in 67 a small sterile tube and analyzed using both an NP150 (size range 50 to 250nm) and 68 NP400 (size range 200 to 1000nm) nanopore at a 45mm stretch. Calibration was 69 performed using CPC200 OR SKP400 calibration particles (Izon) as a standard 70 was verified using Ponceau red solution. Membranes were blocked in 3% BSA-TBS, 1 hours 112 at RT before proceeding to the antibody incubation. All primary antibody incubation was 113 performed in blocking buffer overnight at 4°C. Horseradish peroxidase-conjugated anti-114 mouse or anti-rabbit antibodies were used as secondary antibodies and incubated for 1h at 115 RT. Immunocomplexes were visualized by chemiluminescence using ECL according to 116 manufacturer's instuctions (Pierce, #32106). A G-BOX Imaging System (GeneSys) was used 117 to catch up the specific bands, and the optical density of each band was measured using 118 specific software (GeneTools, Syngene). After initial immunodetection, membranes were 119 stripped of antibodies and re-probed with antibody against total protein or another protein 120 with the same molecular weight. All proteins for each panel were assessed on one 121 membrane, therefore actin expression need to be determined once to control for loading for 122 all these proteins. The difference between the proteins of interest and the control loading 123 protein of the same sample was calculated as relative content and presented graphically. Phospho-ATF2 (#5112) and actin (#8457) were purchased from Cell Signaling Technology 128 (Danvers, MA) and used at the recommended dilution for immunoblotting (1:1000). 129

RNA isolation and quality control. 130
ECFC were harvested from cultures dishes and total RNA was extracted using the mirVana 131 miRNA Isolation Kit (Ambion), according to the manufacturer's recommendations. The 132 quantity and quality of the RNA were assessed using a Nanodrop (Thermo Science, Orsay, 133 France) and a 2100 bioanalyzer (Agilent Technologies, Massy,France), respectively. All 134 samples used in microarrays study, showed common, high quality RNA Integrity Numbers 135 (RIN 9.0-9.8).  was run in duplicate, and the relative fold change was determined using the 2 -CT methods 146 with CT-ECFC as baseline, normalized to RPL13A expression. 147

Microarrays. 148
The microarray study was performed using microarrays chip that included 45,000 probes (1 149 microarray for each sample, 4x44K Whole Genome Microarray G4112F) and the One-Color 150 Microarray-Based Gene Expression Analysis based on the Agilent Technologies 151 procedures 7 . Briefly, 400ng of total RNA were converted to cDNA, followed by in vitro 152 transcription and incorporation of Cy3-CTP into nascent cRNA. The hybridization was 153 performed for 17h at 65°C. cRNA labeling and hybridization performance were performed 154 and all parameters checked were found within the manufacturers specifications. Arrays were 155 scanned as described in the manufacturers' protocol. Signal intensities on 20 bit tiff images 156 were calculated by Feature Extraction software (FE, Version 8.5; Agilent Technologies containing the EMP fractions were diluted in EBM2 and stored at -80°C until subsequent use. 207 The number of resulting EMP was quantified using high-sensitive flow cytometry as 208 described above. The high-speed last wash supernatant was used as control (vehicle). optimal induction of senescence. 217

Proliferation assay 222
The effect of CM or EMP treatment on HUVEC proliferation was assessed by measurement 223 of BrdU incorporation using the Cell proliferation assay Kit according to the manufacturer's 224 instructions (Roche Diagnostic Mannhein Germany). HUVEC were plated at 6,000 cells/wells 225 in gelatin-coated 96-well culture plates in triplicate and were allowed to attach in complete 226 medium. Subsequently, CM ± depleted in EMP or medium containing either vehicle (SN) or 227 CT or PT-EMP was added for 24h. Thereafter, 10µM BrdU was added to each well and 228 incubated overnight. BrdU incorporation was assayed by spectrophotometry using an optical 229 density of 450nm. 230

Cell cycle analysis 231
Cell cycle analysis was conducted by using a propidium iodide-based flow cytometry 232 protocol. HUVEC (35,000 cell/wells in 24-well culture dishes) were incubated for 24hours 233 with CM or EMP from each condition, harvested, fixed with 70% cold ethanol, and stained 234 with propidium iodide. At least 10,000 events were acquired per sample with a Gallios flow 235 cytometer system and analyzed using Kaluza® Analysis software (Beckman Coulter). 236

THP-1 adhesion to HUVEC 237
THP-1 monocytic cells were cultured in RPMI 1640 with 10% heat-inactivated FCS. They 238 were split one to tenth once a week. Adhesion of the monocytic cell line THP-1 to HUVEC 239 was performed as described by Akeson  analyzed for statistical significant differences using either Tukey's (all group compared to 261 each other) or Dunnet's (groups compared to control group) methodology, as appropriate.

262
Statistical significance was accepted at p-value < 0.05.