MSC isolation and culture
Bone marrow were isolated from femurs and tibias of 8 Sprague-Dawley’s rats. MSC were isolated by plastic adhesion, amplified up to 80–90% confluence in a complete medium containing MEM alpha (Biological Industries) supplemented with 10% fetal calf serum (FCS, HyClone) + 1% penicillin streptomycin (Gibco) and 1 ng/mL of bFGF (Peprotech), then pooled at the same ratio for each donor. The cells were frozen in MEMα + 10% Albumin (Vialebex, LFB) + 1% penicillin streptomycin + 10% Dimethylsulfoxide (Sigma Aldrich) and finally preserved in liquid nitrogen. For in vivo experiments, MSC were thawed and seeded at 3000 to 5000 cells/cm2. Medium was changed one day after thawing. When MSC reached reached 90 to 100% of confluence, they were harvested using trypsine-EDTA (Gibco) and suspended in a sterile syringe at a concentration of 2.106 cells/mL of Ringer Lactate solution (Fresinus).
Then, we verified that rat bone marrow MSC fulfill the minimal criteria defining MSC (10). First, their capacity to adhere to plastic and proliferate were confirmed. Rat MSC expressed the classical surface markers CD29, CD73, CD90, and CD105 and were negative for CD11b, CD45, and HLA-DR. Third, these MSC could differentiate into osteoblasts, chondroblasts and adipocytes in vitro. In addition, formulation (Ringer Lactate suspension) and storage (+ 4 °C) have been set up to optimize the best cell viability.
MSC priming
When MSC reached 80% of confluence the medium was changed and replaced by complete medium + 5 ng/mL rat IL-1β (Peprotech) for 24 h. Then, primed MSC were washed twice, harvested and suspended as described above.
Animal Model
All experimental protocols, procedures and endpoints criteria were approved by the institutional animal committee of Paris University (French Ministry of Research number CEEALV/#9498).
A total of 64 Sprague-Dawley male rats (Janvier Labs) weighting 392 ± 5 g, were included in this protocol. The anesthesia was carried out by spontaneous mask ventilation with isoflurane (Iso-Vet, Piramal Healthcare). A subcutaneous injection of buprenorphine (50 µg/kg) and 100 µL of local anesthetic: Lidocaine hydrochloride (Xylocaine, Aguettant) 2.8 mg/mL + Levobupivacaine Chlorydrate (Chirocaine, Abbvie) 1.4 mg/mL were performed on the incision site. These injections were repeated before the animals were awakened at the same dose. Both carotid artery and jugular vein were catheterized for bleeding the animals (controlled hemorrhage) and for administration of fluid resuscitation, treatment and blood re-transfusion. After surgery, an injection of 300 µL heparin Choay was carried out (100 UI/mL, Sanofi). Further details are provided in the supplemental data – Methods.
Experimental protocol
46 Rats were allocated to four groups as described below:
Sham Group
Rats underwent surgery without bleeding, fluid resuscitation and blood re-transfusion. They received 0.9% saline (2 mL/h) after anesthesia.
Hemorrhage Shock Group (HS)
Blood was withdrawn to reach a 35 ± 2 mmHg MAP and maintained for 90 minutes. At T0H, Ringer Lactate infusion started (2 mL/kg/min) equal to twice the volume of spoiled blood. After 1 hour (T1H), blood withdrawan in heparinized (25 UI) syringes was re-transfused.
Treated Groups with naive MSC (MSCn) or IL-1β-primed MSC (MSCp): At T0H after hemorrhagic shock, MCSs were slowly injected intravenously into the flow of Ringer Lactate (106 cells). Blood withdrawal was re-transfused at T1H.
For all animals, the catheters were removed at T2H, the skin sutured and then the isoflurane was stopped. At the time of sacrifice (T6H), samples of blood and organs (kidney, liver) were taken.
Plasma analysis
Whole blood collected from the aorta at T6H on heparinized syringe was centrifuged at 2600 g during 12 min. Plasma concentrations of cytokines: IL-1β, IL-6 and IL-10 (Duoset, R&D) as well as Cystatin C (Quantiquine R&D), were measured in duplicates using sandwich ELISA essay according to the manufacturer’s instructions.
Plasma biochemistry analyses were performed using a Cobas 6000 automated analyzer in Percy military hospital (Clamart, France). Liver function was assessed by plasma levels of aspartate aminotransferase (AST), alanine aminotransferase (ALT) and gamma-glutamyl transferase (GGT). Kidney function was assessed by the plasma level of blood urea nitrogen (BUN) as well as the plasma levels of creatinine.
Histological and Immunohistochemistry
The organs were fixed in 4% buffered formaldehyde, and embedded in paraffin. Sections were then cut into 4 µm slices for Hematoxylin-Eosin-Safran (HES) (liver) and Periodic acid-Schiff (PAS) (kidney) staining. Slides were scanned and examined on CaseViewer software by two blind investigators. Kidney Tubular injury score (tubular dilation, tubular cast formation and tubular necrosis) (16) was quantified in the external medulla. Liver injury score (congestion, vacuolization and necrosis of hepatocytes) (17), was quantified on 30 randomly selected fields. For kidney injury molecule-1 (KIM-1), sections were labeled with goat anti-rat KIM-1 antibody at 1 µg/mL concentration (R&D). Slides were scanned and 5 randomly selected fields of the kidney external medulla were analysis with Fiji software. Further details are provided in the supplemental data – Methods.
Cytometry analysis
For phenotype characterization of white blood cells, a multi-parameter analysis was carried out by flow cytometry (Navios, Beckman Coulter®). Whole blood was collected in tube field with cellular antigen stabilization solution (Transfix, Cliniscience) and kept at 4 °C up to 14 days. Red blood cells were lysed with Macs lysing solution. To minimize non-specific antibody binding, after centrifugation, cells were incubated in a buffer containing PBS, 2% human albumin (LFB) and 2 µg/mL polyvalent human immunoglobulin (R&D Systems) before staining. Cells were then incubated with fluorescent monoclonal antibodies: CD8a (OX8, eBioscience), CD4 (OX35, eBioscience), CD3 (1F4, BioRad), CD80 (3H5, BioRad), CD279 or PD-1 (KLH, Bioss), CD274 or PD-L1 (KLH, Bioss), CD11b/c (REA, Miltenyi), CD28 (REA, Miltenyi), CD45 5REA, Miltenyi), CD86 (REA, Miltenyi) and MHC2 (REA, Miltenyi), at a saturating concentration for 20 minutes at + 4 °C. Isotype antibodies, non-stain cells and fluorescence minus one were used as controls. The FlowJo software was used to set up gating and analyze positivity frequencies of the makers of interest.
Statistical analyses
All data are presented as median ± interquartile range and statistical analyses were performed on the Prism 7 Graphpad software. Mann-Whitney statistical analysis was used to test the effect of our hemorrhaging model versus Sham. Dunn’s test (adjust for multiplicity) were used to compare the effects of the treatments (MSCn and MSCp) to the HS group. The significance level was defined as a p value < 0.05.