Implanted Adipose-Derived Stem Cells Attenuate Small-for-Size Liver Graft Injury by Secretion of VEGF in Rats

Authors

  • T. Ma,

    1. Department of Hepatobiliary and Pancreatic Surgery, the First Affiliated Hospital, School of Medicine, Zhejiang University, Key Laboratory of Multi-Organ Transplantation of Ministry of Public Health, Hangzhou, China
    Search for more papers by this author
  • H. Liu,

    1. Department of Hepatobiliary and Pancreatic Surgery, the First Affiliated Hospital, School of Medicine, Zhejiang University, Key Laboratory of Multi-Organ Transplantation of Ministry of Public Health, Hangzhou, China
    Search for more papers by this author
  • W. Chen,

    1. Department of Hepatobiliary and Pancreatic Surgery, the First Affiliated Hospital, School of Medicine, Zhejiang University, Key Laboratory of Multi-Organ Transplantation of Ministry of Public Health, Hangzhou, China
    Search for more papers by this author
  • X. Xia,

    1. Department of Hepatobiliary and Pancreatic Surgery, the First Affiliated Hospital, School of Medicine, Zhejiang University, Key Laboratory of Multi-Organ Transplantation of Ministry of Public Health, Hangzhou, China
    Search for more papers by this author
  • X. Bai,

    1. Department of Hepatobiliary and Pancreatic Surgery, the First Affiliated Hospital, School of Medicine, Zhejiang University, Key Laboratory of Multi-Organ Transplantation of Ministry of Public Health, Hangzhou, China
    Search for more papers by this author
  • L. Liang,

    1. Department of Hepatobiliary and Pancreatic Surgery, the First Affiliated Hospital, School of Medicine, Zhejiang University, Key Laboratory of Multi-Organ Transplantation of Ministry of Public Health, Hangzhou, China
    Search for more papers by this author
  • Y. Zhang,

    1. Department of Emergency Medicine, the First Affiliated Hospital, School of Medicine, Zhejiang University, Key Laboratory of Multi-Organ Transplantation of Ministry of Public Health, Hangzhou, China
    Search for more papers by this author
  • T. Liang

    Corresponding author
    1. Department of Hepatobiliary and Pancreatic Surgery, the First Affiliated Hospital, School of Medicine, Zhejiang University, Key Laboratory of Multi-Organ Transplantation of Ministry of Public Health, Hangzhou, China
      Tingbo Liang, liangtingbo@zju.edu.cn
    Search for more papers by this author

Tingbo Liang, liangtingbo@zju.edu.cn

Abstract

Graft injury after small-for-size liver transplantation impairs graft function and threatens the survival of the recipients. The use of adipose-derived stem cells (ADSCs) for liver injury protection and repair is promising. Our aim was to investigate the role of vascular endothelial growth factor (VEGF) secreted by ADSCs in the treatment of small-for-size liver graft injury. Studies were performed using ADSCs with VEGF secretion blocked by RNA interference. In vitro, ADSCs prevented apoptosis of freshly isolated liver sinusoidal endothelial cells (LSECs) by secretion of VEGF. Syngeneic 35% orthotopic liver transplantation followed by implantation of syngeneic ADSCs through the portal vein system was performed using Wistar rats. We found VEGF secreted by implanted ADSCs improved graft microcirculatory disturbances, serum liver function parameters and survival. The improved microcirculatory status was also reflected by reduced hepatocellular damage, especially LSEC apoptosis and improved liver regeneration. These effects were accompanied by decreased expression of endothelin receptor type A, increased Bcl-2/Bax ratio, decreased expression of Bad and elevated proportion of phosphorylated Bad. In conclusion, implanted syngeneic ADSCs attenuated small-for-size liver graft injuries and subsequently enhanced liver regeneration in a rat 35% liver transplantation model. The VEGF secreted by implanted ADSCs played a crucial role in this process.

Abbreviations: 
ADSCs, adipose derived stem cells; ALT, alanine aminotransferase; AST, aspartate aminotransferase; Bad, Bcl-2 associated death promoter; Bax, Bcl-2-associated X protein; Bcl-2, B-cell lymphoma 2; BM-MSCs, bone marrow derived mesenchymal stem cells; ET-1, endothelin-1; ETAR, endothelin receptor type A; GFP, green fluorescence protein; HBSS, Hank's balanced salt solution; HGF, hepatocyte growth factor; IFM, intravital fluorescence microscopy; IOD, integral optical density; IRI, ischemia-reperfusion injury; LSECs, liver sinusoidal endothelial cells; PCNA, proliferating cell nuclear antigen; SFS, small-for-size; SFSLT, small-for-size liver transplantation; SFSS, small-for-size syndrome; TEM, transmission electron microscope.

 

Introduction

Living donor liver transplantation is increasingly used globally because of shortages of whole liver grafts. Many have reported that small-for-size (SFS) liver grafts (graft-to-recipient weight ratio of <0.8–1% or graft volume/standard liver volume of <30–40%) affect graft function and endanger recipients by inducing small-for-size syndrome (SFSS; Refs. 1–4). Work on SFSS prevention has focused on reducing portal hypertension and protecting grafts from ischemia-reperfusion injury (IRI) early postsurgery (5–7). Our previous study showed that SFS liver grafts are subjected to not only acute IRI but also prolonged microcirculatory disturbance, a key factor in exacerbating graft injury and inducing graft failure (8), indicating that the search for new approaches to protect SFS liver grafts should be a priority.

The past 10 years have seen increasingly rapid advances in the field of adult stromal stem cell research. They are considered to hold considerable promise for the future of stem cell therapy for various diseases. Recently, it was reported that implanted bone marrow derived mesenchymal stem cells (BM-MSCs) can inhibit hepatocellular apoptosis and stimulate liver regeneration in a rat liver IRI model (9), indicating that adult stromal stem cells may benefit SFS liver graft recovery. As an easily accessible, abundant and reliable stem cell source, adult stromal stem cells isolated from adipose tissue or adipose-derived stem cells (ADSCs), represent a promising candidate for stem cell–based therapy. With their well-known capacities of multilineage transdifferentiation and rapid proliferation, ADSCs have also been shown to secrete significant quantities of angiogenic and antiapoptotic factors, including vascular endothelial growth factor (VEGF) and hepatocyte growth factor (HGF; Ref. 10). Many studies have shown ADSCs to attenuate microcirculatory disturbances and revascularize ischemic tissues by promoting reperfusion, limiting apoptosis and facilitating angiogenesis in ischemic tissues by secreting growth factors (10–13), directly supporting the emerging concept that local paracrine factors are key to ADSC-based therapy. That ADSCs can improve the myocardium postinfarction by a paracrine mechanism without myocardial differentiation further emphasizes the therapeutic importance of soluble factors from these cells (14).

On the basis of the observations above, one might hypothesize that implanted ADSCs could be useful in SFS liver graft salvage by attenuating the microcirculatory disturbances that play a predominant role in SFS liver graft injury (8). Here, we explored the therapeutic potential of syngeneic ADSCs on SFS liver transplantation recipients in a rat SFS liver transplantation model. The contribution of VEGF secreted by ADSCs in SFS liver graft protection was also determined by RNA interference (RNAi)-mediated silencing of VEGF.

Materials and Methods

Animals

Male Wistar rats (200–250 g, 8-week-old) from the Shanghai Laboratory Animal Center, Chinese Academy Sciences were housed in a controlled 12-h light/dark cycle environment with access to chow and water ad libitum. All animal experiments were performed according to the Zhejiang University guidelines for animal care and approved by the Animal Ethics Review Committees of Zhejiang University.

Cell isolation and identification

ADSCs were isolated and cultured essentially as described before (15). In brief, aseptically recovered inguinal adipose tissue from Wistar rats was digested enzymatically (37°C, 1 h) in 0.075% type I collagenase (Gibco, Grand Island, NY, USA). After centrifugation, the stromal cell fraction was filtered and cultured in Dulbecco's Modified Eagle's medium (DMEM, Gibco) containing 10% fetal bovine serum (FBS, Gibco) and antibiotics. Passage 3 cells were verified by immunostaining of surface markers for analysis by flow cytometry (Beckman Coulter, Brea, CA, USA) and immunofluorescence microscopy (Olympus, Tokyo, Japan) using fluorescent-conjugated antibodies against rat CD14, CD34, CD45, CD90 (Caltag Laboratories, San Francisco, CA, USA), CD29 (Biolegend, San Diego, CA, USA), CD31, CD68 (Abcam, Cambridge, MA, USA), CD133 (Novus Biologicals, Littleton, CO, USA) and respective isotypes. ADSCs were induced to differentiate into adipocytes and osteocytes to confirm their multipotency according to published methods (15). Oil Red O and Alizarin Red stainings were performed to detect the lipid droplets and matrix mineralization in induced cells, respectively.

Liver sinusoidal endothelial cells (LSECs) were isolated using a perfusion method as described before with minor modifications (16). In brief, the rat liver was perfused in vivo and digested ex vivo with 0.05% collagenase IV (Gibco). Obtained cells were filtered and LSECs enriched by centrifugation for 20 min at 900×g with a two-step percoll gradient (Invitrogen, Carlsbad, CA, USA). The intermediate zone between 25% and 50% percoll was suspended in culture medium containing 2% FBS, antibiotics, 2 mM l-glutamine in DMEM. After selective adherence of Kupffer cells, cells were plated on collagen I coated 6-well plates for 4 h and extensively washed for further coculture experiments.

Transduction of lentiviral RNAi vectors

Lentiviral vectors (GenePharma, Shanghai, China) carrying either a shRNA oligonuleotide sequence (5’-TGGGCCTCTGAAACCATGAATTCAAG
AGATTCATGGTTTCAGAGGCCCTTTTTTC-3’) targeted to VEGF mRNA (VEGFi-shRNA) or a scramble shRNA sequence (5’-TGTTCTCCG
AACGTGTCACGTTTCAAGAGAACGTGACACGTTCGGAGAACTTTTTTC-3’) with no homology to any rat gene were used. Both vectors also express green fluorescence protein (GFP), the latter serving as a negative control (NC-shRNA). ADSCs were transduced with lentiviral constructs as described before (12). In brief, passage 1 cells were seeded at 104 cells/cm2 in 6-well plates (Corning, Corning, NY, USA). Lentivirus at 5, 25 or 100 multiplicity of infection (MOI) and 5 μg/mL polybrene were added to DMEM with 10% FBS in 2 mL total volume. After 24 h, the medium was replaced for expansion of transduced cells. Transduction efficiency was analyzed by fluorescence microscopy and flow cytometry for GFP expression. Populations of transduced ADSCs containing >95% GFP-positive cells were used in the studies described later. The VEGF concentrations in the supernatants of transduced or nontransduced ADSCs were determined by enzyme-linked immunosorbent assays (ELISA) using a rat VEGF ELISA Kit (R&D Systems, Minneapolis, MN, USA).

Coculture assay and detection of apoptosis by flow cytometry

A 6-well transwell system (0.4 μm pore, Corning) was used for the coculture experiment. Freshly isolated LSECs were seeded in the lower chamber (collagen I coated) at 1 × 106 per mL and incubated at 37°C to allow cell adhesion. After 4 h, the medium in the lower chamber was refreshed and media containing 1 × 104 or 3 × 105 NC-shRNA ADSCs, 1 × 104 or 3 × 105 VEGFi-shRNA ADSCs or 1.5 mL fresh medium without cells (control) were added to the upper chamber to establish a separated coculture system. After 24 h, LSECs in the lower chamber were extracted and apoptosis was examined by flow cytometry using an Annexin V/PI Apoptosis Detection Kit (Merck Chemicals, Darmstadt, Germany).

Experimental design and SFS liver transplantation

Both SFS liver transplantation and ADSCs implantation were performed using syngeneic Wistar rats. Recipient rat groups received orthotopic SFS liver transplantation: (i) without cell implantation (control) (ii) with NC-shRNA ADSCs and (iii) with VEGFi-shRNA ADSCs. Each group consisted of 30 liver transplants. Liver tissue and blood of six recipients used for intravital fluorescence microscopy (IFM) were sampled postoperatively on days 1, 3 and 7. Another 12 rats in each group were used for the survival study.

Rats were fasted overnight before the operation. After general anesthesia with sodium pentobarbital (30–50 mg/kg, i.p.), the caudate lobes, left lateral lobe, right superior and inferior lobes of the rat liver were resected by sutures and ligations, with the median lobe spared to obtain a 35% liver graft. The excised liver graft was preserved in 4°C normal saline for about 35 min (35.2 ± 1.9 min) before implantation. Orthotopic nonarterialized partial liver transplantations were performed using the two-cuff technique with minor modifications (17). Immediately after graft reperfusion, 2 × 106 syngeneic ADSCs suspended in 1 mL Hank's balanced salt solution (HBSS) were injected into a terminal branch of the superior mesenteric vein of the recipient using a 28G insulin needle over 5 min. One milliliter HBSS was injected into the same vein of control group recipients. After the operation, rats recovered spontaneously without further treatment and >90% of the rats survived surgery.

IFM and quantitative analysis of liver microcirculation

This procedure was performed according to our published method (8). In brief, rats were anesthetized, injected with dextran-fluorescein (MW: 4,000; 2 μmol/kg i.v.; Invitrogen) to enhance contrast. Liver microcirculation was viewed and images captured under an inverted fluorescence Olympus IX81 microscope (Olympus). For each rat, 10–15 hepatic acini were recorded.

Images were analyzed using Image-Pro Plus 6.0 (Media Cybernetics, Silver Spring, MD, USA). As described before (18), lobules were graded into three categories (4× magnification): not perfused (no staining), irregularly perfused (patchy staining) and well perfused (homogeneous staining). A lobular perfusion index was calculated using the formula (Nw + 0.5 Ni)/Nt, where Nw and Ni (perfused area of lobule <50%) are the number of well-perfused and irregularly perfused lobules, respectively and Nt refers to the total number of lobules analyzed. Sinusoidal diameters were analyzed by measuring diameters of eight sinusoids at the mid-zonal regions of randomly selected perfused acini at 20× magnification.

Real-time PCR

Total mRNA was extracted from frozen tissue samples by Trizol (Invitrogen) and reverse transcribed using an RT reagent kit with gDNA eraser (TaKaRa, Shiga, Japan). Real-time PCR was performed using an ABI 7,500 Sequence Detection System (Applied Biosystems, Foster City, CA, USA) with SYBR-green I (TaKaRa). Relative gene expression profiles were determined by normalizing to the housekeeping gene (β-actin) using the 2−ΔCt method (19). The primer sequences for the various genes are as follows: endothelin-1 (ET-1) forward: ACCTGGACATCATCTGGGTCAAC and reverse: TTTGGTGAGCACACTGGCATC; endothelin receptor type A (ETAR) forward: TCTCTGCGCTCTCAGTGTGGA and reverse: AGCCGATTGCTTCTGGGATG; B-cell lymphoma 2 (Bcl-2) forward: TGAACCGGCATCTGCACAC and reverse: CGTCTTCAGAGACAGCCAGGAG; Bcl-2-associated X protein (Bax) forward: AGACACCTGAGCTGACCTTGGAG and reverse: GTTGAAGTTGCCATCAGCAAACA; Bcl-2 associated death promoter (Bad) forward: GGCAGCCAATAACAGTCATCA and reverse: GGTACGAACTGTGGCGACTC; β-actin forward: GGAGATTACTGCCCTGGCTCCTA and reverse: GACTCATCGTACTCCTGCTTGCTG.

Western blot

Western blot was used to analyze the protein expression in liver tissues. Antibodies against ET-1 and ETAR were from Santa Cruz Biotechnology (Santa Cruz, CA, USA). Antibodies against Bcl-2, Bax, Bad and phosphorylated Bad (p-Bad) were from Cell Signaling Technology (Danvers, MA, USA). The relative amount of each protein was normalized to GAPDH expression (Cell Signaling Technology) and all antibodies were used at 1:1000 dilution.

Terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) assay

An In Situ Cell Death Detection Kit, POD (Roche Molecular Biochemicals, Indianapolis, IN, USA) was used to detect cellular apoptosis in liver sections according to the manufacturer's instructions. Cells with brown stained nuclei were considered apoptotic cells.

Immunohistological analysis

Paraffin-embedded liver sections were stained for Proliferating Cell Nuclear Antigen (PCNA; Santa Cruz) and Factor VIII (Epitomics, Burlingame, CA, USA). The percentages of PCNA-positive nuclei were determined under a light microscope and the positive stainings of Factor VIII were analyzed using integral optical density (IOD) values measured using Image-Pro Plus 6.0 software (Media Cybernetics). The IOD of three independent positively stained microscopic fields per section and two sections per liver graft were calculated. Fields containing large vascular structures (portal/central venules and hepatic arterioles) were excluded and only microvessels were analyzed.

Statistical analysis

Data were analyzed with SPSS 11.0 software (SPSS, Chicago, IL, USA). One-way analysis of variance (ANOVA) was used to compare group variables, followed by least-significant difference (LSD) posthoc testing when indicated. One-way ANOVA with Games–Howell was performed for distributions where equal variances could not be assumed. Values were expressed as means ± standard deviation. A p-value <0.05 was considered significant. Kaplan–Meier survival analysis was performed and differences between survival curves were estimated using the log-rank test.

Results

Identification of ADSCs and transduction efficiency of lentiviral vectors

The isolated ADSCs expanded rapidly in culture and could differentiate into both adipocytes and osteocytes (Figure 1C). Most cells expressed CD29 and CD90 (over 95%), whereas few expressed CD14 (0.3%), CD31 (0.7%), CD34 (0.4%), CD45 (0.6%), CD68 (0.8%) and CD133 (0.4%; Figure 1A and B), confirming that the cells isolated from rat inguinal fat pads were ADSCs as reported before (20) and that contamination from other cell sources was limited.

Figure 1.

Identification and purity of ADSCs. Flow cytometry revealed that the distribution of ADSCs that stained for CD14, CD31, CD34, CD45, CD68 and CD133 (shaded regions) did not differ from that of the isotype control (open regions). The majority of cells positively stained for CD29 and CD90 (shaded regions) compared with the isotype control cells (open regions) (A). Immunofluorescence analysis confirmed that ADSCs stained positively for CD29 and CD90, but not CD14, CD31, CD34, CD45, CD68 and CD133 (B). ADSCs resembled fibroblasts in culture and were successfully induced into adipocytes and osteocytes (C).

Most transduced ADSCs were shown to express GFP (>95%) by fluorescence microscopy and flow cytometry (Figure 2A and B). The VEGF level in supernatants of VEGFi-shRNA ADSCs (20.1 ± 13.1 pg/105 cells) was significantly lower than those of NC-shRNA ADSCs (90.4 ± 16.9 pg/105 cells) and normal control ADSCs (82.7 ± 13.8 pg/105 cells, Figure 2C, **p < 0.01), suggesting VEGF expression was efficiently blocked in VEGFi-shRNA ADSCs.

Figure 2.

Effect of VEGF secretion or its inhibition by RNAi from ADSCs on LSECs in coculture assay. The Transduction efficiencies of indicated lentiviral RNAi vectors were confirmed by fluorescence microscopy (A) and flow cytometry for GFP (B, open regions represent normal control cells). VEGF expression and secretion into the supernatant were significantly blocked by VEGFi-shRNA Transduction of ADSCs (C, **p < 0.01). The percentage of LSECs undergoing apoptosis (D, upper left region) was significantly lower when cocultured with 3 × 105 NC-shRNA ADSCs than with 1 × 104 NC-shRNA ADSCs, 1 × 104 or 3 × 105 VEGFi-shRNA ADSCs and control (B, **p < 0.01).

Antiapoptotic effect of ADSCs on LSECsin vitro

Most of the freshly isolated LSECs did not survive beyond 72 h without additional vascular growth factors (data not shown). The proportion of apoptotic LSECs was much lower when cocultured with 3 × 105 NC-shRNA ADSCs (7.3 ± 6.2%) than with control (42.9 ± 8.9%), 1 × 104 NC-shRNA ADSCs (42.7 ± 4.9%), 1 × 104 (41.3 ± 5.4%) or 3 × 105 (39 ± 7.1%) VEGFi-shRNA ADSCs (Figure 2D), suggesting ADSCs could prevent apoptosis of cultured LSECs by secretion of VEGF.

Improved liver function, survival and liver microhistology of small-for-size liver transplantation (SFSLT) recipients by implanted NC-shRNA ADSCs

The serum alanine aminotransferase (ALT) and aspartate aminotransferase (AST) levels from rats of the NC-shRNA group were significantly lower than those from the control (ALT: day 3, 56.8 ± 16.2 U/L vs. 141.1 ± 67.9 U/L, p < 0.05; day 7, 40 ± 20.9 U/L vs. 94.7 ± 37.7 U/L, p < 0.05. AST: day 3, 189.7 ± 120.6 U/L vs. 634.2 ± 279.9 U/L, p < 0.05; day 7, 248.8 ± 145.3 U/L vs. 473.2 ± 225.3 U/L, p < 0.05.; Figure 3A) and VEGFi-shRNA groups postsurgery (ALT: day 1, 333 ± 102.2 U/L vs. 592 ± 174.9 U/L, p < 0.05; day 3, 56.8 ± 16.2 U/L vs. 128.7 ± 55.6 U/L, p < 0.05; day 7, 40 ± 20.9 U/L vs. 111.5 ± 33.7 U/L, p < 0.01. AST: day 1, 882.7 ± 129.1 U/L vs. 1186.3 ± 104.5 U/L, p < 0.01; day 3, 189.7 ± 120.6 U/L vs. 415.7 ± 186.2 U/L, p < 0.05; day 7, 248.8 ± 145.3 U/L vs. 586.7 ± 160.5 U/L, p < 0.01; Figure 3A). Nine of twelve rats (75%) in the NC-shRNA group survived >100 days after SFSLT, compared to the significantly lower rate of 33% (4/12) for both control and VEGFi-shRNA groups (Figure 3B). The general clinical and pathological manifestations of dying recipients in all of these groups were similar. They were progressively weak and jaundiced, with ascites and intestinal swelling in the abdomen, but without obstruction of the inflow and outflow blood vessels and the common bile duct. These results indicated that the implanted NC-shRNA ADSCs promoted liver function recovery, rescuing the recipients from acute liver failure and the VEGF production by these ADSCs was crucial to these protective effects.

Figure 3.

Effect of VEGF expression in implanted ADSCs on rat liver functions and survival. Recipients from the NC-shRNA group showed significantly lower serum ALT and AST levels (A, **p < 0.01 and *p < 0.05, respectively) and higher survival rate compared with the other two groups (B, *p < 0.05 vs. control group; †p < 0.05 vs. VEGFi-shRNA group).

Microcirculatory disturbances, portal inflammatory cell infiltration and hepatocellular necrosis were notable in postoperative liver sections from control and VEGFi-shRNA groups. Normal liver architecture was nearly restored in liver sections from the NC-shRNA group (Figure 4H, H&E staining). Under the transmission electron microscope (TEM), nuclear hyperchromatosis and shrinkage of sinusoidal endothelial cells, irregular large gaps between sinusoidal lining cells and mitochondria swelling of hepatocytes were apparent in samples from control and VEGFi-shRNA groups (Figure 4, TEM). By contrast, normal ultrastructure of sinusoidal endothelial cells and hepatocytes from the NC-shRNA group were nearly restored.

Figure 4.

Effect of VEGF expression from implanted ADSCs on liver histological manifestations of rats. H&E: Grafts from control and VEGFi-shRNA groups showed sinusoidal dilation/disturbance, red blood cell deposition and hepatocellular acidophilic degeneration (postoperative day 1); focal hepatocellular necrosis and portal lymphocytic infiltration (postoperative day 3); diffuse necrosis and disruption of lobular architecture (postoperative day 7). Normal liver architecture was nearly restored in grafts from NC-shRNA group postoperatively. TEM: nuclear hyperchromatosis and shrinkage of sinusoidal endothelial cells (arrows), irregular large gaps between sinusoidal lining cells (arrowheads) and mitochondria swelling of hepatocytes (asterisks) were apparent in samples from the control and VEGFi-shRNA groups. By contrast, normal ultrastructure of sinusoidal endothelial cells and hepatocytes from the NC-shRNA group were nearly restored.

Improved hepatic microcirculation of SFSLT grafts by implanted NC-shRNA ADSCs

Microcirculatory disturbances were predominant in control and VEGFi-shRNA groups, characterized by extravasation of erythrocytes and petechial bleedings and cessation of individual sinusoid blood flow postoperatively. By contrast, the microcirculatory status in the NC-shRNA group was much improved compared to the other two groups with no apparent microcirculatory disturbance visible (Figure 5A). Quantitatively, microcirculatory injuries in control and VEGFi-shRNA groups showed significantly lower lobular perfusion indexes (control vs. NC-shRNA: day 1, 0.78 ± 0.02 vs. 0.91 ± 0.02, p < 0.01; day 3, 0.70 ± 0.04 vs. 0.84 ± 0.01, p < 0.01. VEGFi-shRNA vs. NC-shRNA: day 1, 0.79 ± 0.02 vs. 0.91 ± 0.02, p < 0.01; day 3, 0.72 ± 0.03 vs. 0.84 ± 0.01, p < 0.01; Figure 5B) and larger sinusoidal diameters (control vs. NC-shRNA: day 1, 8.71 ± 0.61 μm vs. 6.60 ± 0.52 μm, p < 0.01; day 3, 7.20 ± 0.27 μm vs. 6.35 ± 0.34 μm, p < 0.05. VEGFi-shRNA vs. NC-shRNA: day 1, 8.87 ± 0.15 μm vs. 6.60 ± 0.52 μm, p < 0.01; day 3, 7.33 ± 0.38 μm vs. 6.35 ± 0.34 μm, p < 0.05; Figure 5C) than in the NC-shRNA group. Real-time PCR showed ETAR expression in the NC-shRNA group was significantly lower than in control and VEGFi-shRNA groups (0.0023 ± 0.0005 vs. 0.0051 ± 0.0016 and 0.0045 ± 0.001, respectively, p < 0.05; Figure 5D), with no significant difference in ET-1 expression among these three groups (control: 0.00049 ± 0.0003, NC-shRNA: 0.00055 ± 0.00043, VEGFi-shRNA: 0.00042 ± 0.0002, p > 0.05; Figure 5D). These results were further confirmed at the protein level by Western blot.

Figure 5.

Effect of VEGF expression from implanted ADSCs on microcirculatory manifestations of rats. Hepatic microcirculation in control and VEGFi-shRNA groups was characterized by extravasation of erythrocytes/petechial bleedings (arrows) or cessation of individual sinusoid blood flow (arrowheads) with nearly normal microcirculation status in the NC-shRNA group (A). Graft microcirculatory disturbances in control and VEGFi-shRNA groups were quantified by lowered lobular perfusion index and dilated sinusoidal diameters (B, C, **p < 0.01 and *p < 0.05). Real-time PCR and Western blot showed the expression of ET-1 in these groups did not differ from each other, whereas the mRNA and protein levels of ETAR were significantly lower in the NC-shRNA group (D, *p < 0.05).

Attenuation of apoptosis in vivo by implanted NC-shRNA ADSCs

Results of the TUNEL assays showed that the proportion of apoptotic cells in the NC-shRNA group was significantly lower than in control and VEGFi-shRNA groups (0.007 ± 0.01 vs. 0.07 ± 0.01 and 0.08 ± 0.008, respectively, p < 0.01,Figure 6A). Real-time PCR showed that the Bcl-2/Bax ratio in the NC-shRNA group was significantly higher than in the control group (0.21 ± 0.04 vs. 0.11 ± 0.01, p < 0.05; Figure 6B). Bad mRNA expression in the NC-shRNA group was significantly lower than control and VEGFi-shRNA groups (0.0031 ± 0.0004 vs. 0.0046 ± 0.0008 and 0.0044 ± 0.0005, respectively, p < 0.05; Figure 6B). These results were confirmed at the protein level and phosphorylated Bad was more notable in the NC-shRNA group. Taken together, implanted NC-shRNA ADSCs prevented cell apoptosis in vivo, accompanied by down-regulation and phosphorylation of Bad and less notably, up-regulation of Bcl-2/Bax ratio.

Figure 6.

Detection of hepatocellular apoptosis in vivo after implantation of ADSCs. Apoptotic cells, indicated by brown stained nuclei (arrows), showed a scattered distribution pattern in liver sections from control and VEGFi-shRNA groups (postoperative day 1), whereas few positive staining was visible in those from the NC-shRNA group (A, **p < 0.01). Real-time PCR and Western blot analyses showed that the Bcl-2/Bax ratio was much higher in the NC-shRNA group compared with the control group. Bad expression in the NC-shRNA group was significantly lower than the other two groups and phosphorylated Bad was more notable in the NC-shRNA group (B, **p < 0.01 and *p < 0.05).

Improved hepatocyte regeneration and microvascular regeneration in grafts with implanted NC-shRNA ADSCs

Most cells were PCNA-positive in immunohistochemically stained sections from the NC-shRNA group on postoperative day 3 (77.5 ± 7.4%) and significantly fewer PCNA-positive cells were in the control (26.3 ± 7.5%) and VEGFi-shRNA (28.4 ± 3.6%) groups (p < 0.01; Figure 7). Factor VIII staining showed many more positively stained cells in the NC-shRNA group than control and VEGFi-shRNA groups on postoperative day 7 (1.7×105± 5.2×104 vs. 6.5×104± 2.8×104 and 5.6×104± 0.9×104, respectively, p < 0.01; Figure 7). It was notable that Factor VIII-positive cells were well arranged in the NC-shRNA group, whereas they were mostly clustered or scattered in the other two groups. These results showed implanted NC-shRNA ADSCs promoted hepatocyte regeneration and microvascular regeneration.

Figure 7.

Hepatocellular and microvascular regeneration after implantation of ADSCs. Enhanced hepatocellular regeneration (PCNA staining) and microvascular regeneration (Factor VIII staining) were apparent in grafts from the NC-shRNA group, compared with the control and VEGFi-shRNA groups (**p < 0.01). Factor VIII-positive cells were well-arranged in the NC-shRNA group but were mostly clustered or scattered in the other two groups.

Discussion

In clinical practice, the partial liver graft size is strictly limited to over ∼40% of the ideal liver weight to avoid graft failure, which is lethal to the recipient without retransplantation (2,3,21). In our model, a large majority of rats died early after 35% liver transplantation, although rats subjected to 70% partial hepatectomy have been shown to recover well and even 90% hepatectomy is not lethal (22). A possible explanation is that the partial liver graft suffers not only ischemia-reperfusion injury but also mechanical injuries related to hemodynamic force (23,24). After reperfusion, excessive blood flow (relative to graft size) induces shear stress, causing damage to LSECs and activation of kupffer cells, which both contribute to acute liver failure (2,23,25–27). The manifestations of microcirculatory disturbances and subsequent graft injuries in this 35% liver transplantation model were well demonstrated in our previous paper (8). In this study, the general and histopathological manifestations of dying recipients indicated acute liver failure after transplantation.

As a model for adult stromal stem cell-based therapy for SFS liver graft injury, we first tested BM-MSCs in SFSLT in a pilot study, but without promising results (data not shown). Although BM-MSCs transduced to express HGF can promote SFS liver graft recovery (28), the risks of viral agents are challenging for clinical applications. Compared with BM-MSCs, the effects of ADSCs seem to be superior in salvaging ischemic tissue (11). Therefore, we chose to explore the possible salutary effects of implanted ADSCs in SFS liver graft recovery.

As reported by Zuk et al. (15), pluripotent ADSCs are enriched by plastic adherence of the stromal-vascular cell fraction isolated from adipose tissue. Although most of the contaminating cells were eliminated or extremely reduced after several passages in our study, it was essential for us to rule out the effects of other types of cells, especially those that may secrete VEGF in large amounts. In this study, the predominantly negative stainings of CD14, CD31, CD34, CD45, CD68 and CD133 indicated that contamination with hematopoietic stem cells, endothelial cells, fibrocytes and macrophages was extremely low.

VEGF is regarded as a vascular endothelial mitogen (29) and demonstrated as a survival factor for endothelial cells (30). As concluded by some (31,32), liver regeneration is, at least partly, an angiogenesis-dependent phenomenon and the endothelial cell is a key mediator of adult tissue mass regeneration in this partial hepatectomy model. Delivery of VEGF can increase liver mass in mice but do not stimulate growth of hepatocytes in vitro, unless LSECs are also present, indicating that VEGF promotes liver sinusoidal reconstruction, angiogenesis and liver regeneration by stimulating sinusoidal endothelial cell proliferation (33). VEGF treatment has also been shown to significantly reduce the mortality rate of acute liver failure in rats through maintenance of LSEC architecture and antiapoptotic effects (34). Recently, it was reported that VEGF-dependent regenerative stimuli are important in inducing proliferation of mature hepatocytes throughout the residual liver after partial hepatectomy (35). Hence, it could conceivably be hypothesized that VEGF secreted by implanted ADSCs might promote the recovery and regeneration of SFSLT grafts by protecting LSECs and by subsequently enhancing the regeneration of LSECs and hepatocytes. In this study, grafts of the NC-shRNA group maintained both an enhanced liver regeneration rate and microvascular regeneration rate, consistent with improved liver function and increased survival rate. It seems possible that these results were because of enhanced regenerative stimuli by VEGF secreted by implanted ADSCs and of the attenuation of microcirculatory disturbances.

In this study, the implanted ADSCs were shown to attenuate microcirculatory disturbances after SFS liver transplantation. The improved microcirculatory status was also reflected by reduced hepatocellular damage, especially LSEC apoptosis and improved liver function parameters. Gene microarray profiles of SFSLT grafts have revealed that genes associated with hemodynamic disturbances such as ET-1 and ETAR are predominantly up-regulated and proapoptotic genes including Bad are also up-regulated postoperatively (36). In another study, SFSLT is shown to lead to increased ET-1 and ETAR expression and ETAR blockade reduces SFS graft injuries by attenuating microcirculatory disturbances and reducing hepatocellular damage (27), indicating that ETAR may be a potentially valuable therapeutic target for SFSS. Here, we found that NC-shRNA ADSCs significantly down-regulated ETAR expression but not ET-1 expression. Thus, the reduced ETAR levels, observed in this study, may have benefited the SFS grafts. However, more research is required to clarify the association between ETAR and SFS graft injury and how VEGF may interact with ETAR should be illustrated in future studies. We also found that the VEGF secreted by ADSCs prevented the apoptosis of LSECs and hepatocytes, accompanied by increased Bcl-2/Bax ratio, reduced Bad expression and elevated proportion of phosphorylated Bad. The present findings are consistent with other studies which demonstrated that VEGF inhibits cell apoptosis along with up-regulation of Bcl-2 (37) and down-regulation of Bad (38).

As demonstrated by others, the endothelial-derived angiocrine factors (e.g. HGF) that promote liver regeneration are induced by VEGF (35). And up-regulated expression of VEGF in SFS liver grafts might facilitate the activities of macrophages (39), indicating the VEGF secreted by implanted ADSCs may facilitate the production of endogenous VEGF, as activated macrophages secrete VEGF. Therefore, the VEGF produced by implanted ADSCs might act by autocrine/paracrine pathways to stimulate cells to release various proangiogenic and antiapoptotic factors in SFS liver grafts, all of which may contribute to the improved graft injuries. However, more work needs to be done before drawing this conclusion in the SFS liver transplantation model.

In summary, implanted ADSCs protected SFS liver grafts from microcirculatory disturbances, facilitating liver function recovery, regeneration of hepatocytes and endothelial cells and survival of recipients. We showed that VEGF played a crucial role in the salutary effects of ADSCs in this rodent model. However, further studies are needed in various animal models to verify these conclusions and to investigate other roles (e.g. immunomodulatory effect) of implanted ADSCs in SFS liver grafts to promote their development for clinical application.

Acknowledgments

This study was sponsored by the Key Program of Natural Science Foundation of Zhejiang Province, China (No.Z2080283), the National Science Fund for Distinguished Young Scholars (No. 30925033) and the Zhejiang Provincial Program for the Cultivation of High-level Innovative Health Talents. The authors would like to acknowledge Dr. Wei Ding for assistance with immunohistological stainings, A/Prof. Filip Braet for assistance with isolation of LSECs and Elixigen Corporation for proofreading the manuscript.

Disclosure

The authors of this manuscript have no conflicts of interest to disclose as described by the American Journal of Transplantation.

Ancillary