Potential conflict of interest: Nothing to report.
Endothelin-1 receptor A blocker darusentan decreases hepatic changes and improves liver repopulation after cell transplantation in rats
Article first published online: 16 JAN 2014
© 2014 by the American Association for the Study of Liver Diseases
Volume 59, Issue 3, pages 1107–1117, March 2014
How to Cite
Bahde, R., Kapoor, S., Viswanathan, P., Spiegel, H.-U. and Gupta, S. (2014), Endothelin-1 receptor A blocker darusentan decreases hepatic changes and improves liver repopulation after cell transplantation in rats. Hepatology, 59: 1107–1117. doi: 10.1002/hep.26766
Supported, in part, by National Institutes of Health grants R01 DK088561, R01 DK071111, and P30-DK41296. R.B. received a postdoctoral fellowship from the German Research Foundation (DFG).
- Issue published online: 25 FEB 2014
- Article first published online: 16 JAN 2014
- Accepted manuscript online: 1 OCT 2013 01:34PM EST
- Manuscript Accepted: 19 SEP 2013
- Manuscript Received: 16 JUL 2013
Cell transplantation-induced hepatic ischemia and recruitment of vasoconstrictors (e.g., endothelin-1; Edn1) leads to clearance of transplanted cells and poses problems for liver repopulation. Therefore, we determined whether darusentan (DAR), which potently blocks Edn1 receptor type A, could benefit cell engraftment. We transplanted primary F344 rat hepatocytes with or without DAR in dipeptidyl peptidase IV–deficient rats. Analysis of microcirculatory events included hepatic ischemia, endothelial injury, including with gene expression arrays, and activations of Kupffer cells (KCs), neutrophils, or hepatic stellate cells (HSCs). The retrorsine-partial hepatectomy model was used for liver repopulation studies. Whether DAR was directly cytoprotective was examined in cultured rat hepatocytes or CFSC-8B rat HSCs. We found that DAR induced hepatic sinusoidal vasodilation, caused more transplanted cells to be deposited in liver parenchyma, and decreased hepatic ischemia and endothelial injury. This lessened perturbations in expression of endothelial biology genes, including regulators of vessel tone, inflammation, cell adhesion, or cell damage, versus drug-untreated controls. Moreover, in DAR-treated animals, cell transplantation-induced activation of KCs, albeit not of neutrophils, decreased, and fewer HSCs expressed desmin. In DAR-treated rats, improvements in cell engraftment led to greater extent of liver repopulation, compared to drug-untreated controls. In cell-culture assays, DAR did not stimulate release of cytoprotective factors, such as vascular endothelial growth factor, from HSCs. Moreover, DAR did not protect hepatocytes from tumor necrosis factor alpha– or oxidative stress–induced toxicity. Endothelin receptor A blockade in vitro did not improve engraftment of subsequently transplanted hepatocytes. Conclusion: Systemic administration of DAR decreases hepatic ischemia-related events and thus indirectly improves cell engraftment and liver repopulation. This vascular mechanism may permit the development of combinatorial drug-based regimens to help optimize cell therapy. (Hepatology 2014;59:1107–1117)
Dulbecco's modified Eagle's medium
dipeptidyl peptidase IV deficient
endothelin type A/type B
hepatic stellate cell
liver sinusoidal endothelial cell
methyl thiazol tetrazolium
National Institutes of Health
tumor necrosis factor alpha
vascular endothelial growth factor
Efficient engraftment of transplanted cells in the liver was apparent early on as a barrier for cell therapy in humans.[1, 2] Cell engraftment requires depositing of cells in liver sinusoids, which causes hepatic ischemia, tissue injury, and inflammation as a result of vaso-occlusion, and 80%-90% transplanted cells are lost within 1-2 days. This cell clearance is mediated, in part, by cytokines, chemokines, and receptors activated by neutrophils, Kupffer cells (KCs), liver sinusoidal endothelial cells (LSECs), or hepatic stellate cells (HSCs)[3-5] and, in part, by instant blood-mediated reaction involving procoagulant activity and complement. The underlying mechanisms are complex because endothelial damage, without thrombotic occlusion, simultaneously allows transplanted cells to enter liver parenchyma,[7, 8] whereas release by HSC of vascular endothelial growth factor (VEGF), matrix metalloproteinases, and so on, protect transplanted cells and facilitate parenchymal remodeling during cell engraftment. However, on balance, cell transplantation-induced microcirculatory alterations are deleterious and must be overcome. For instance, direct-acting vasodilators (i.e., nitroglycerine, phentolamine, or prostacyclin) improve cell engraftment.[3, 9] Use of such drugs to control harmful microcirculatory events will be highly significant for cell therapy.
Recently, endothelin-1 (Edn1), a potent vasoconstrictor that transduces its effects through type A (Ednra) or type B (Ednrb) receptors, was incriminated in cell transplantation-induced changes. Bosentan, a nonspecific blocker of Ednra/Ednrb, improved cell engraftment, emphasizing the role of Edn1. However, in bosentan recipients, transplanted cells did not proliferate or repopulate the liver. Whether this was the result of displacement by bosentan of harmful ligands that might have produced changes in naïve transplanted cells was possible, e.g., plasma Edn1 levels were elevated in Edn1 receptor knockout mice. This possibility was confirmed when hepatic Edn receptors were blocked beforehand by bosentan in vitro, because transplanted cells could then proliferate and repopulate the liver. Although intracellular signaling from Edn1 receptors is ill-defined, this includes compensatory and/or opposing effects. Of Edn1 receptors, selective blockade of Ednra is considered desirable, because Ednrb may be cytoprotective. Therefore, Ednra blockers were developed, for example, darusentan (DAR), which is in the late clinical phase for vascular conditions and shows promise for liver conditions.[14, 15] Here, we considered that Ednra blockade with DAR will improve cell transplantation-induced microcirculatory changes and thereby cell engraftment. We performed cell transplantation assays in dipeptidyl peptidase IV deficient (DPPIV−) F344 rats, including the retrorsine/partial hepatectomy (PH) model of liver repopulation.[3-5, 7-9]
Materials and Methods
DAR and Chemicals
Unless specified, all reagents and chemicals were from Sigma-Aldrich Chemical Co. (St. Louis, MO). DAR (Knoll, Ludwigshafen, Germany) was dissolved to 10 mg/mL of normal saline containing 0.24 mL of 1 N of NaOH for pH 11 with final pH to 7.5 with 0.1 N of HCl.
Six- to eight-week-old DPPIV− rats, weighing 120-150 g, were from the Special Animal Core of Marion Bessin Liver Research Center (Bronx, NY). Donor F344 rats were from the National Cancer Institute (Bethesda, MD). Rats were anesthetized with ketamine and xylazine. The animal care and use committee at Albert Einstein College of Medicine (Bronx, NY) approved protocols per guidelines from the National Institutes of Health (NIH; Bethesda, MD).
Hepatocytes were isolated by collagenase liver perfusion, as described previously (3). For vascular effects of DAR, 2 × 106 15-µm latex microspheres (New England Nuclear, Boston, MA) were injected intrasplenically (IS) 2 hours after 2.5-10.0 mg/kg of DAR or vehicle given intraperitoneally (IP). After 2 hours, rats were killed for microsphere distributions in liver. Right ventricular (RV) pressures were measured by PE50 cannulae in the internal jugular vein before and after DAR. For hepatocyte transplantation, DAR or vehicle was given IP 2 hours before IS injection of 2 × 107 freshly isolated hepatocytes. For KC activity, 0.1-mL carbon particles encriched from India ink (Pelikan no. 17; Pelikan Hannover, Germany) were injected IS 6 hours after cells and rats were killed 30 minutes later. To study the effects of repeatedly administered DAR, 2 × 107 hepatocytes were transplanted twice 10 days apart. For liver repopulation, 30 mg/kg of retrorsine were given IP to 6- and 8 week-old DPPIV− rats, followed 4 weeks later by two-thirds PH, when 5 × 106 hepatocytes were injected IS.
Messenger RNA Studies
Total RNA was extracted by Trizol Reagent (Invitrogen, Carlsbad, CA). RNA was reverse-transcribed by the RT PCR Array First Strand Kit, and for reverse-transcription polymerase chain reaction, we used RT Real-Time SYBR Green PCR master mix (SABiosciences, Frederick, MD). Rat endothelial cell biology array with 84 gene probes was used (PARN-015; SABiosciences; http://www.sabiosciences.com/rt_pcr_product/HTML/PARN-015A.html), according to the manufacturer's instructions. We analyzed replicate samples from 3 rats per condition. The 2-ΔΔCt method was used for fold differences. Data were first normalized with β-actin in individual samples. Gene expression differences of >1.5-fold with P < 0.05 were considered significant.
Studies With Cells In Vitro
All conditions were in triplicate and studies were repeated at least twice. For cytotoxicity assays, 1 × 105 hepatocytes were cultured per cm2 plastic dishes in Dulbecco's modified Eagle's medium (DMEM) with 10% fetal bovine serum and antibiotics. For tumor necrosis factor alpha (TNF-α) studies, cells were preincubated with 200 ng/mL of actinomycin D for 2 hours before 10 ng/mL of TNF-α (Serologicals Corp., Atlanta, GA) for 16-18 hours. For oxidative stress (OS), hepatocytes were cultured for 16-18 hours and exposed to 2 mM of H2O2 for 90 minutes. DAR (in 1-100-μM doses) was added immediately after cell plating. Rat CFSC-8B HSCs were cultured in DMEM. Culture conditions included normoxia (21% O2 and 5% CO2) or hypoxia (5% O2, 5% CO2, and 11% N2). Conditioned medium from CFSC-8B cells was obtained in serum-free DMEM after 16-18 hours. Cell viability was analyzed by thiazolyl blue (methyl thiazol tetrazolium; MTT) utilization. For additional transplantation studies, cells were incubated for 1 hour on ice in serum-free DMEM or DMEM containing 10 μM of DAR.
Multiple samples were analyzed from at least 3-4 rats per condition. Five-micrometer sections were prepared from tissues frozen in methylbutane at −80°C. For gamma-glutamyl transpeptidase (GGT), sections were fixed in chloroform-acetone and reacted with N-γ-L-glutamyl-4-methoxy-β-naphthylamide (Polysciences, Inc., Washington, PA). GGT+ areas were measured in 25-50 fields centered on portal radicles. KCs containing carbon were counted in 50 periportal areas per rat under 200× microscopic magnification. Neutrophils with myeloperoxidase (MPO) activity were localized by peroxidase indicator reagent (Sigma-Aldrich), as previously described. MPO+ cells were counted in 50 areas under 200× microscopic magnification. Sections were stained with desmin antibody (Dako Cytomation, Carpinteria, CA), as previously described. Desmin+ HSCs were counted under 400×. For DPPIV histochemistry, sections were fixed in chloroform-acetone (1:1, vol/vol) with toluidine blue counterstain, as previously described.[3-5, 7-9] DPPIV+ transplanted cells were counted in 100 areas centered on portal radicles under 100× microscopic magnification. After transplanting cells twice, engraftment was analyzed 14 days after the final cell transplantation session. For liver repopulation, 3 weeks after cell transplantation, DPPIV+ cells were analyzed in 25 areas per rat by ImageJ software (NIH, Bethesda, MD).
Hyaluronic Acid and VEGF Measurements
Serum was assayed for hyaluronic acid (HA) and culture medium for VEGF by kits (Corgenix, Inc., Westminster, CO; and Rat VEGF Quantikine ELISA, RRV00; R&D Systems, Minneapolis, MN), as previously described. Serum alanine aminotransferase (ALT) and bilirubin were analyzed by an automated clinical system.
All studies included 3-8 rats per condition. Major findings were reproduced at least twice. Data are expressed as means ± standard deviation. Significances were analyzed by t tests, Mann-Whitney's rank-sum tests, or analysis of variance, including Holm-Sidak's pair-wise comparisons by SigmaStat 3.1 statistical software (Systat Inc., Point Richmond, CA). P < 0.05 was considered significant.
Rats tolerated DAR without morbidity. Administration of DAR (in doses of 2.5-10.0 mg/kg) had no effect on RV pressures. We observed no gross liver abnormalities, histological changes, or differences in serum ALT or bilirubin levels after DAR (not shown).
DAR Improved Cell Engraftment by Altering Sinusoidal Distribution of Transplanted Cells
Dose-ranging studies showed that more latex microspheres were in the distal portions of hepatic sinusoids after 10 mg/kg versus either 1 or 5 mg/kg of DAR (not shown). Consequently, we chose 10 mg/kg of DAR for studies. We observed more transplanted cells in liver sinusoids after DAR (P < 0.05; Fig. 1). Large numbers of transplanted cells (∼90%) were cleared after 6 hours or later in controls as well as DAR-treated rats, with similar kinetics of transplanted cell clearance. However, in DAR-treated rats, more cells engrafted, including after 6 hours, 3 days, and 7 days (P < 0.05). Cumulatively, transplanted cell numbers were 1.5- to 2.5-fold higher in DAR-treated rats (P < 0.05). Greater entry of transplanted cells in sinusoids was reflected by transplanted cells in fewer portal vein radicles after 6 hours (mean of 31% in DAR-treated versus 69% in control rats after 6 hours; P < 0.05). Because transplanted cell numbers in liver were similar after 3 and 7 days, this indicated that cells surviving after initial clearance had engrafted.
Effects of DAR on Cell Transplantation-Induced Changes
Expression of endothelial biology genes altered the most 6 hours after cell transplantation. In controls or DAR-treated rats, expression was unaltered in 18 of 84 (21%) genes, including vascular genes (e.g., angiotensin receptor 1b, Edn2, Ednra, and so on). By contrast, expression was different by 1.5- to 225-fold in 50 of 84 (60%) genes (P < 0.05; Table 1). These genes included regulators of vessel tone, angiogenic factors, inflammatory cytokines, chemokines, and receptors, cell death or apoptosis, cell adhesion, and so on. In DAR-treated rats, expression of 17 genes up-regulated by cell transplantation decreased, including regulators of vessel tone or angiogenesis (Adam17, Fgf1, Pgf, and transforming growth factor beta 1), apoptosis (Casp1), inflammation (Ccl2, Cx3cl1, Il1b, and Tnfsf10), cell adhesion (Pecam, Sell), or thrombosis (Plg, Thbd). As an exception, Il-6 was up-regulated in DAR-treated rats. Expression in DAR-treated rats of several inflammatory cytokines (Cxcl1, Cxcl2, Il11, Il7, and Tnf) was similar to controls.
|Gene Description||Gene Symbol||Fold Up or Down in Cells Alone Versus Untreated Controls||Fold Up or Down in Cells + DAR Versus Untreated Controls||Fold Difference Cells + DAR Versus Cells Alonea||Comment|
|Angiotensin I converting enzyme (peptidyl-dipeptidase A) 1||Ace||3.2||2.5||0.8||No effect|
|A disintegrin and metalloproteinase domain 17 (tumor necrosis factor, alpha, converting enzyme)||Adam17||1.4||−2.0||0.4||Down-regulation|
|Angiotensinogen (serpin peptidase inhibitor, clade A, member 8)||Agt||2.0||1.5||0.7||No effect|
|Angiopoietin 1||Angpt1||−1.4||−2.1||0.7||No effect|
|Annexin A5||Anxa5||1.7||1.5||0.9||No effect|
|Bcl2-associated X protein||Bax||1.8||1.4||0.8||No effect|
|B-cell leukemia/lymphoma 2||Bcl2||2.0||1.8||0.9||No effect|
|Bcl2-like 1||Bcl2l1||1.8||−1.2||0.6||No effect|
|Baculoviral IAP repeat-containing 1b||Birc1b||−1.5||−2.8||0.5||Down-regulation|
|Burkitt lymphoma receptor 1||Blr1||4.8||−2.2||0.7||No effect|
|Caspase 3, apoptosis-related cysteine protease||Casp3||1.7||1.2||0.7||No effect|
|Caspase 6||Casp6||1.6||1.2||0.7||No effect|
|Chemokine (C-C motif) ligand 2||Ccl2||9.4||4.3||0.5||Down-regulation|
|Cadherin 5 (predicted)||Cdh5_predicted||1.1||−1.2||0.3||Down-regulation|
|CASP8- and FADD-like apoptosis regulator||Cflar||2.0||1.3||0.6||No effect|
|Procollagen, type XVIII, alpha 1||Col18a1||1.6||1.1||0.6||No effect|
|CASP2 and RIPK1 domain-containing adaptor with death domain (predicted)||Cradd||−1.1||−2.2||0.5||Down-regulation|
|Colony-stimulating factor 2 (granulocyte-macrophage)||Csf2||13.5||15.6||1.2||No effect|
|Chemokine (C-X3-C motif) ligand 1||Cx3cl1||3.8||1.1||0.3||Down-regulation|
|Chemokine (C-X-C motif) ligand 1||Cxcl1||181.0||130.4||0.7||No effect|
|Chemokine (C-X-C motif) ligand 2||Cxcl2||224.9||267.5||1.2||No effect|
|Endothelial cell growth factor 1 (platelet-derived)||Ecgf1||−4.0||−5.6||0.7||No effect|
|Endothelin 1||Edn1||5.4||6.3||1.2||No effect|
|Fibroblast growth factor 1||Fgf1||−1.3||−5.5||0.2||Down-regulation|
|FMS-like tyrosine kinase 1||Flt1||−1.2||−2.0||0.6||No effect|
|Fibronectin 1||Fn1||2.0||−1.3||0.6||No effect|
|Intercellular adhesion molecule 1||Icam1||4.8||3.2||0.7||No effect|
|Integrin alpha 5||Itga5||−1.1||−1.9||0.6||No effect|
|Integrin beta 1 (fibronectin receptor beta)||Itgb1||1.9||−1.3||0.6||No effect|
|Integrin beta 3||Itgb3||1.7||1.2||0.7||No effect|
|Kinase insert domain protein receptor||Kdr||−2.6||−3.9||0.7||No effect|
|V-kit Hardy-Zuckerman 4 feline sarcoma viral oncogene homolog||Kit||−1.3||−1.9||0.7||No effect|
|Matrix metallopeptidase 9||Mmp9||23.2||14.6||0.6||No effect|
|Nitric oxide synthase 2, inducible||Nos2||55.6||39.5||0.7||No effect|
|Natriuretic peptide receptor 1||Npr1||−1.4||−1.6||0.9||No effect|
|Platelet-derived growth factor receptor, alpha polypeptide||Pdgfra||1.9||−1.3||0.6||No effect|
|Platelet/endothelial cell adhesion molecule||Pecam||1.2||−1.5||0.4||Down-regulation|
|Chemokine (C-X-C motif) ligand 4||Cxcl4||2.2||2.3||1.1||No effect|
|Placental growth factor||Pgf||13.6||7.1||0.5||Down-regulation|
|Plasminogen activator, tissue||Plat||2.1||1.9||0.9||No effect|
|Plasminogen activator, urokinase||Plau||2.2||2.4||1.1||No effect|
|Prostaglandin I2 (prostacyclin) synthase||Ptgis||5.8||3.6||0.6||No effect|
|Inositol 1,3,4,5,6-pentakisphosphate 2-kinase||Ippk||2.0||1.6||0.8||No effect|
|Ras homolog gene family, member B||Rhob||8.9||5.5||0.6||No effect|
|Selectin, platelet||Selp||40.5||31.7||0.8||No effect|
|Serine (or cysteine) peptidase inhibitor, clade E, member 1||Serpine1||181.0||144.0||0.8||No effect|
|Endothelial-specific receptor tyrosine kinase||Tek||−1.8||−2.3||0.8||No effect|
|Transforming growth factor, beta 1||Tgfb1||1.8||−1.9||0.5||Down-regulation|
|Thrombospondin 1||Thbs1||19.1||18.9||1.0||No effect|
|Tissue inhibitor of metallopeptidase 1||Timp1||19.5||19.8||1.0||No effect|
|Tumor necrosis factor (TNF superfamily, member 2)||Tnf||9.0||8.3||0.9||No effect|
|Fas (TNF receptor superfamily, member 6)||Fas||2.1||1.6||0.8||No effect|
|Tumor necrosis factor (ligand) superfamily, member 10||Tnfsf10||−1.3||−3.0||0.4||Down-regulation|
|Fas ligand (TNF superfamily, member 6)||Faslg||−1.2||−1.5||0.8||No effect|
|TNFAIP3 interacting protein 2||Tnip2||3.1||2.1||0.7||No effect|
|Vascular cell adhesion molecule 1||Vcam1||2.2||1.3||0.6||No effect|
|Von Willebrand factor||Vwf||−1.2||−1.7||0.7||No effect|
Tissue analysis confirmed hepatic ischemia after cell transplantation. In controls, hepatic GGT was expressed, carbon uptake increased in KC, and MPO+ neutrophils appeared after 6 hours; also, desmin+ HSCs appeared after 3 days (Fig. 2). Serum HA levels in rats 6 hours after cell transplantation were higher (70 ± 8 versus 35 ± 6 ng/mL in healthy rats; P < 0.05), confirming endothelial injury. By contrast, in DAR-treated rats, hepatic ischemia and related events decreased because GGT expression declined by 50% (P < 0.05), numbers of KCs with carbon declined by 23% (P < 0.05), and serum HA levels decreased to 55 ± 5 ng/mL (P < 0.05); also, 40% fewer desmin+ HSCs were observed (P < 0.05). However, the number of MPO+ neutrophils was unchanged in DAR-treated rats.
These tissue findings corresponded to gene expression differences in DAR-treated rats. Less GGT expression and KC or HSC activity were reflected in lower expression of genes regulating vessel tone, angiogenesis, cell adhesion, thrombosis, apoptosis, or macrophage/monocyte-associated cytokines (Ccl2 and Cx3cl1), whereas expression of other cytokines (e.g., Tnf, Il6, Cxcl1, and Cxcl2) likely reflected continued neutrophil activation.
Repeated Hepatocyte Transplantation and DAR
Rats tolerated multiple doses of DAR without morbidity. In drug-untreated controls, 2 weeks after two doses of 2 × 107 cells, transplanted cell numbers increased by 2-fold versus single dose of 2 × 107 cells (P < 0.05; Fig. 3). In DAR-treated rats, transplanted cell numbers increased by 4- and 2-fold versus one or two doses of cells in controls, respectively (P < 0.05). This indicated that DAR was repeatedly beneficial for cell engraftment.
DAR and Transplanted Cell Proliferation
We used retrorsine/PH-preconditioned rats for liver repopulation with cells transplanted immediately after isolation or after incubation with DAR or vehicle in vitro (Fig. 4A). Engraftment of cells transplanted immediately was similar in DAR-treated retrorsine/PH-preconditioned rats and DAR-treated nonpreconditioned rats (P = not significant). In retrorsine/PH-preconditioned rats without DAR, 7% ± 2% of liver was repopulated after 3 weeks. In DAR-treated retrorsine/PH-preconditioned rats, 4.4-fold more liver was repopulated (31% ± 13%; P < 0.05; Fig. 4B). Therefore, DAR did not interfere with hepatocyte proliferation. By contrast, transplantation of cells incubated in vitro produced different results. First, engraftment of DAR-incubated cells was only 33% ± 6% of vehicle-treated cells; second, proliferation of vehicle-incubated cells was inferior to cells transplanted immediately after isolation; and third, proliferation of DAR-incubated cells was worse than vehicle-incubated cells (P < 0.05) in each case (Fig. 4C).
Effects of DAR on Cells In Vitro
Cell viability was unchanged after culture of hepatocytes with vehicle or 1-100 μM of DAR. Hepatocyte viability declined by approximately 50% after culture with TNF-α or H2O2 (P < 0.05; Fig. 5). In hepatocytes cultured with 1-100 µM of DAR plus TNF-α or H2O2, viability neither improved nor worsened.
To determine whether DAR could have affected release of paracrine factors from HSCs, which benefited cell engraftment in the case of bosentan, we studied rat CFSC-8 HSCs. Culture of these cells under hypoxia increased VEGF secretion into medium (22 ± 4 vs. 44 ± 10 pg/mL; P < 0.05). The conditioned medium protected hepatocytes from TNF-α toxicity. Culture of CFSC-8 cells with DAR under hypoxia led to secretion of only 12 ± 7 pg/mL of VEGF (P < 0.05); moreover, this medium did not additionally protect hepatocytes from TNF-α. Therefore, HSCs did not likely contribute to cell engraftment from release of hepatoprotective paracrine factors in DAR-treated rats.
We established that DAR improved transplanted cell engraftment without interfering with transplanted cell proliferation and liver repopulation. These benefits were realized with relatively small doses of DAR. Previously, DAR was found to be safe for cardiovascular applications in much higher doses, for example, 50 mg/kg daily (in rats) and up to approximately 6 mg/kg or more daily in humans over extended periods.[16, 17] Even 2-to 3-fold greater initial cell engraftment has enormous benefit on the kinetics of liver repopulation, as shown here and in previous studies.[3-5, 7-9] The mechanisms by which DAR was helpful involved direct effects on microcirculation, rather than indirect effects on either transplanted cells or native cells (e.g., HSCs). Therefore, systemic administration will be necessary of DAR for benefits.
In DAR-treated rats, cell transplantation-induced vascular alterations improved, including at gene expression and tissue levels, although not all perturbations were resolved (e.g., inflammation continued). After cell transplantation, hepatic ischemia occurs immediately, but other changes evolve over hours, as shown here. After cell transplantation, hepatic blood flow was restored by sinusoidal vasodilatation, as documented with live video microscopy in nitroglycerine-treated rats. Similarly, directly acting drugs, such as prostacyclin, and the alpha-adrenergic blocker, phentolamine, decreased cell transplantation-induced vasoconstriction. Several classes of vascular drugs were ineffective (e.g., β-blockers, calcium-channel blockers, or angiotensin-converting enzyme inhibitors) because these drugs dilated large vessels, but not hepatic sinusoids.[3, 9] In the case of DAR, hepatic sinusoidal vasodilatation occurred, as indicated by entrapment of fewer cells in portal radicles, the presence of more cells in distal sinusoidal locations, and decreases in hepatic GGT expression, which was originally demonstrated to be activated by ischemic occlusion of hepatic sinusoids, and has since been well-established as a convenient reporter of cell transplantation-induced ischemia.[3, 9] Moreover, expression of several genes regulating vessel tone, cell adhesion, endothelial damage, thrombosis, and so on, was decreased in recipients of DAR, consistent with the role of Edn1-mediated vascular events after cell transplantation. It should be noteworthy that DAR blocks Edn1-induced responses in vascular smooth muscle cells, which regulate vessel tone, and also has direct effects on LSECs because endothelial fenestrae are dilated. Similarly, by improving sinusoidal perfusion, DAR decreased hepatic ischemia-reperfusion injury,[14, 15] which will be in agreement with sinusoidal vasodilatation after cell transplantation in DAR-treated rats.
However, the effects of DAR in the cell-transplantation setting differed from other drugs, including nitroglycerine and prostacyclin, and, particularly, bosentan, as summarized in Table 2. Whereas DAR and bosentan both dilated hepatic sinusoids with restoration, at least partly, of endothelial biology gene expression, only DAR decreased hepatic ischemia, endothelial injury, and KC activation, as judged by decreased GGT expression, serum HA levels, and carbon uptake in KCs. These vascular effects of DAR were similar to the effects of nitroglycerine and prostacyclin on hepatic GGT expression, endothelial injury (lower serum HA levels), and activation of KC after cell transplantation. Whether DAR will modulate procoagulant activity of cells after transplantation for further benefits is unknown.
|Process||Effects of DAR||Effects of Bosentana||Comment|
|Hepatic sinusoidal vasculature||Vasodilation||Vasodilation||Concordant effect|
|Hepatic ischemia (GGT expression)||Decreased||No change||Discordant effect|
|LSEC injury (increased serum HA levels)||Decreased||No change||Discordant effect|
|Endothelial biology gene expression||Decreases in regulators of vessel tone, cell damage, adhesion, thrombosis||Decreases in regulators of vessel tone, cell damage, adhesion, thrombosis||Concordant effect|
|Activation of neutrophils||No change||No change||Concordant effect|
|Macrophage (KC) response||Decreased||No change||Discordant effect|
|Activation of HSCs||Desmin expression decreased; no stimulation of VEGF release||Desmin expression increased; VEGF release stimulated||Discordant effect|
|Transplanted cell engraftment||Increased||Increased||Concordant effect|
|Benefits for repeated cell transplantation||Yes||Unknown||Needs further study|
|Transplanted cell proliferation||No interference||Interference||Discordant effect|
|In vitro drug manipulation of cellsbefore transplantation||No benefit||Beneficial||Discordant effect|
Of note, neither DAR nor bosentan prevented cell transplantation-induced neutrophil activation. This inability to inhibit neutrophil activation was shared by nitroglycerine and prostacyclin. Continuing expression of several inflammatory cytokines and chemokines, including Il6, Tnf, and so on, which is associated with activated neutrophils, despite DAR (or other vascular drugs), was in agreement with this finding. Previously, Edn1 was incriminated in transcriptional regulation of inflammatory cytokine expression by Ednra, along with Ednrb,[21, 22] because these receptors share intracellular signaling pathways. Displacement of Edn1 from its receptors by DAR or bosentan could certainly have amplified such intracellular inflammatory perturbations.
Differences also emerged in cell transplantation-induced activation of HSCs. DAR decreased desmin expression in HSCs in vivo without inducing VEGF release in vitro, whereas bosentan and nitroglycerine increased desmin expression in HSCs in vivo and bosentan induced VEGF release in vitro. This difference was significant because release of VEGF and other hepatoprotective substances was beneficial for cell engraftment.
Other differences concerned roles of DAR and bosentan in protecting hepatocytes from cytotoxic injury. After Ednra blockade by DAR in vitro, we did not observe cytoprotection, whereas after Ednra/Ednrb blockade by bosentan in vitro, cells resisted cytotoxic insults. Similarly, treatment of CFSC-8B cells with DAR did not elicit paracrine benefits by VEGF release, which was different from bosentan treatment of CFSC-8B cells. While bosentan interfered with transplanted cell proliferation, DAR did not interfere with proliferation of transplanted cells.
These differences in the effects of DAR and bosentan provided insights into the role of Edn1 and Ednra/Ednrb in cell transplantation. For cell therapy strategies, incubating cells with bosentan for its cytoprotective effect, along with administration of DAR to animals for hepatic sinusoidal dilatation, should be helpful. This combined approach should be effective for cell transplantation on multiple occasions. Meanwhile, other mechanisms, for example, to inhibit expression of deleterious inflammatory cytokines, promote endothelial permeability without thrombosis, manipulate extracellular matrix components, and induce release of cytoprotective factors from HSC, should help achieve further gains in cell engraftment and liver repopulation.
The general clinical safety of DAR over the long term has been established.[13, 17] Longer half-life of DAR (approximately 24 hours), compared with short-acting vasodilators (e.g., nitroglycerine, prostacylin, and so on[3, 9]), will be advantageous, because after cell transplantation, hepatic vasoconstriction persisted for up to 24 hours. Single doses of DAR needed for cell transplantation should be without toxicity.
Ms. Chaoying Zhang provided technical assistance.