Mesenchymal stromal cells reset the scatter factor system and cytokine network in experimental kidney transplantation

HGF serum levels significantly decreased in grafted rats that were not injected with MSC compared to control normal rats. MSC treatment returned serum HGF to normal levels (Figure 3, upper panel).

Both HGF protein and HGF mRNA were reduced in grafted kidneys compared to native organs. MSC injection prevented the loss of HGF protein (Figure 3, middle panel), although HGF mRNA levels were lower probably for a feedback mechanism (Figure 3, lower panel).

The widespread expression of Met protein in tubules of native kidneys (Figure 4, panel A) was almost completely abrogated in untreated grafted kidneys (Figure 4, panel B). MSC infusion prevented Met loss (Figure 4, panel C). The result was confirmed by Western Blot for Met performed on kidney tissue (Figure 4, lower panel).

Immunohistochemistry showed that MSP was expressed in tubular cells of native kidneys (Figure 5 panel a) and it was undetectable in untreated grafts except for some tubules in which it was distributed with luminal pattern (Figure 5, panel b). MSC restored tubular MSP in transplanted rats, but in a different pattern from normal, i.e. luminal instead of cytoplasmic (Figure 5, panel c). As MSP also its receptor, RON, was diffusely expressed with cytoplasmic pattern in tubules of healthy rats (Figure 5, panel d), while in transplanted rats with acute rejection tubular RON expression was significantly decreased (Figure 5, panel e) and reappeared in tubules of allografts treated with MSC (Figure 5, panel f). In untreated grafts RON was expressed also by some inflammatory cells in the interstitium, but it was absent in infiltrating cells of MSC treated rats (Figure 5, panels e and f).

RT-PCR showed a significant decrease of MSP mRNA in the untreated graft. MSC injection was associated with recovery of MSP mRNA expression (Figure 6).

We performed in vitro experiments aimed to understand whether MSC are possible producers of MSP.

PBMC known to be constitutive MSP producers (44) were used as controls. The expression of MSPmRNA was investigated by quantitative PCR in PBMC and in MSC. The expression of MSP mRNA in PBMC was held as the arbitrary unit. MSC expressed a greater amount of constitutive MSP compared with PBMC (Figure 7).

3 μm thick sections cut along the sagittal plane of formalin-fixed kidney were stained with periodic acid-Schiff, and evaluated by two investigators in double blind fashion, using an Olympus IX8 microscope connected with a CCD camera and software imaging analysis Cell-R (Olympus America, Center Valley, PA, USA).

We scored for primary necrosis, i.e. necrosis that was not associated with tubulitis, 40 nonoverlapping high-power fields in 5 (HPF) not consecutive renal sections for each animal and quantified tubular necrosis as the ratio between necrotic tubules number and total tubules counted. A tubule was counted as having necrosis when tubular cells had one or more of the following features: isometric vacuolization, cell membrane rupture with loss of cytoplasm, nuclear fragmentation, detachment from basal membrane. Tubular cell proliferation was evaluated counting cells expressing Proliferating Cell Nuclear Antigen (PCNA). Renal expression of PCNA, MSP, RON, Met, ED1 antigen, Foxp3 and Green Fluorescent Protein (GFP) were studied by immunohistochemistry in formalin fixed tissue in 10 non-overlapping high-power fields of each section. The sections of paraffin embedded tissue were collected on poly-L-lysine-coated slides (Dako, Carpinteria, CA, USA), they were dewaxed in xylol, passed in a decreasing series of alcohol, and finally rehydrated with distilled water. Endogenous peroxidase was blocked with H₂O₂ 3.7% vol/vol followed by H₂O for 15 min. After 3 washings in 150 mM of PBS the sections underwent microwave antigen retrieval. Subsequently they were exposed overnight at 4° C to the following antibodies: 1) monoclonal mouse anti-PCNA antibody (Ab) (Santa Cruz Biotechnology, Santa Cruz, CA, USA), 1:200; 2) polyclonal goat anti-mouse MSP Ab (Santa Cruz Biotechnology), 1:600; 3) monoclonal mouse anti human RON Ab (Transduction Laboratories, Lexington, KY, USA), 1:750; 4) monoclonal mouse anti-human Met Ab (Novocastra Laboratories Ltd, Newcastle, UK), 1:20, 5) ED1 antigen, 1:80 (Serotec LtD, Oxford, UK), 6) anti Foxp3 Ab, 1:50 (eBioscience Ltd, Hatfield, UK), 7) monoclonal mouse anti-GFP antibody IgG1, diluted 1:1000 (Chemicon International, Temecula, CA, USA). After 3 washings in PBS the immunocomplex was visualized with the biotin-streptavidin-peroxidase complex and 3,3-diaminobenzidine (Dako, Glostrup, Denmark). Sections were faintly counterstained with Harris hematoxylin. Negative controls included both omission of the primary Ab and substitution of IgG for primary antibodies. Positive controls for EGFP positive cells were sections of kidney of EGFP rat. We counted the number of PCNA, Foxp3, GFP positive cells, MSP, RON, Met positive tubules number and RON positive interstitial cells in ten renal sections for each animal.

HGF serum levels were measured in control and allografted rats 7 days after transplantation by ELISA (Institute of immunology, Tokyo, Japan). All experiments were quadruplicated.

Tissue samples were frozen at −70° C in liquid nitrogen immediately following nephrectomy. Tissues were pulverized with a Mikro-Dismembrator (B.Braun Biotech International, Melsungen, Germany) in the presence of liquid nitrogen. The powdered whole tissues were washed twice with PBS and homogenized in ice cold buffer containing 10 mM PIPES, pH 6.8, 100 mMol/L NaCl, 5 mMol/L MgCl₂, 300 mMol/L sucrose, 5 mMol/L ethylene-glycosbis-(β–aminoethyl ether)-N-N’-tetraacetic acid (DIM buffer), 1% Triton X-100, 100 mol/L sodium ortho-vanadate, and protease inhibitors (aprotinin 10 g/mL, pepstatin 10 g/mL, leupeptin 50 g/mL, soybean trypsin inhibitor 100 g/mL phenylmethanesulfonyl fluoride 1 mMol/L) (Sigma-Aldrich). Equal amounts of protein (800 g), determined using the BCA Protein Assay Reagent Kit (Pierce Chemical Company), were separated on 7,5% sodium dodecylsulfate-polyacrylamide gel electrophoresis and transferred to nitrocellulose Hybond filters (Amersham - GE Healthcare, Little Chalfont, Buckinghamshire, UK). Filters were probed with goat polyclonal anti HGF α Ab (c-20, Santa Cruz) that recognizes the α subunit of active heterodimer HGF, the rabbit polyclonal anti Met Ab that recognizes pro-Met and mature Met β subunit (c-28, Santa Cruz). The specific binding was detected by the enhanced chemiluminescence system ECL-Plus (Pierce Chemical Company).

Peripheral blood mononuclear cells (PBMC) were isolated from rat blood by standard Ficoll hystopaque density gradient separation (Sigma-Aldrich). Monocytes were collected after 45 min of adherence to plastic culture plates and resuspended in RPMI 1640 medium, FCS 10%, and penicillin/streptomycin 1% (Invitrogen, Carlsbad, CA, USA) at 37° C in humidified 5% CO2 atmosphere.

To understand whether IFN-γ modified monocytes RON expression, we studied RON mRNA and RON protein in circulating monocytes soon after sampling and in monocytes cultured in absence and presence of LPS (10 g/ml, Sigma - Aldrich) and INF-γ (10 g/ml, R&D Systems) for 6 h and for 24 h respectively.

Total RNA was extracted from renal tissue and cell cultures using respectively trizol method and guanidine-based RNeasy Mini Kit (QIAGEN GmbH, Hilden, Germany). All RNA was treated with DNase from RNase-Free DNase Set (QIAGEN) and dissolved in nuclease free water. Extracted RNA was tested for quantity and integrity by spectrophotometric analysis (NanoDrop - Thermo Scientific, Waltham, MA, USA). A total of 1 g of RNA per condition was reverse transcribed into complementary DNA (cDNA) through 1st Strand cDNA Synthesis Kit for Real Time (RT) -PCR (AMV) (Roche Applied Science, Penzber, Germany). cDNA was used to perform RT-PCR analysis in 96-well optical reaction plates, using ABI prism 5700 (Applied Biosystems, Foster City, CA, USA) and the 5-exonuclease assay (TaqMan technology) in a volume of 25 reaction containing TaqMan Universal Master Mix, optimized concentrations of FAM-labelled probe, and specific forward and reverse primers for beta-actin, HGF, MSP and RON selected from Assay on Demand (Applied Biosystems). Controls included RT-PCR with water replacing cDNA. The results were analysed using a comparative method, and the values were normalized to the beta-actin expression and converted into fold change, as previously described [63].

A total of 10 ml of EDTA anticoagulated peripheral blood and 3X10⁵ monocytes were incubated for 30 min with 10μ l of mouse anti rat RON mAb (Transduction Laboratories) diluted 1:10. Subsequently the samples were incubated for 30 min with 2μ l of goat anti mouse PE conjugated followed by incubation for 30 min with 2μ l of mouse anti rat CD11b FITC conjugated (BD Biosciences, San Jos, CA) to identify rat monocytes. Before examination samples of whole blood were treated with lysis buffer. All samples were resuspended in 500μ l of PBS and analyzed in FACScan operating with Cell Quest 3.3 software (BD Biosciences). Goat PE- conjugated IgG antibodies were used as negative controls.

ANOVA followed by the Newman-Keuls test or Kruskal-Wallis test and Student t test were used for comparison of the medians and means. All data was analysed using Graph Pad Prism software.