|ognizant Communication Corporation|
VOLUME 13, NUMBER 6, 2004
Cell Transplantation, Vol. 13, pp. 605-617, 2004
0963-6897/04 $20.00 + 00
Copyright © 2004 Cognizant Comm. Corp.
Printed in the USA. All rights reserved.
Methods of Human Islet Culture for Transplantation
T. B. Murdoch, D. McGhee-Wilson, A. M. J. Shapiro, and J. R. T. Lakey
Clinical Islet Transplant Program, University of Alberta, Edmonton, Alberta, Canada
The ability to maintain isolated human islet preparations in tissue culture has recently been adopted by most islet transplant centers, and improves the safety as well as the practicality of islet transplantation. Maintaining islet viability and recovery, however, remains challenging in a clinical setting, due to stringent conditions required for culture. Islet culture is further complicated by the fact that islets do not form a monolayer. This review aims to clarify media, supplementation, and conditions that have been shown to be relevant to human islets, as well as to offer avenues of future research. Factors examined that may influence islet survival include base medium, glucose concentration, vitamin, inorganic ion, lipid, hormone, growth factor, amino acid, and binding protein composition and concentration, as well as culture temperature and seeding density. In addition, this article reviews novel techniques, such as coculture and matrices, that have been employed in an attempt to improve islet survival and functional viability.
Key words: Islet culture; Supplementation; Islet transplantation; Media; Islet survival
Address correspondence to Jonathan R. T. Lakey, Ph.D., Assistant Professor of Surgery, Clinical Islet Transplant Program, 1074 Dentistry/Pharmacy Building, University of Alberta, Edmonton, Alberta T6G 2N8, Canada. Tel: (780) 492-3077; Fax: (780) 492-6335; E-mail: firstname.lastname@example.org
Comparison of Tolerated and Rejected Islet Grafts: A Gene Expression Study
Tobias Berg,1,2,3 Tao Wu,1,2,4 Brett Levay-Young,2 Neal Heuss,1,2 Yisheng Pan,1,2 Nicole Kirchhof,1,2 David E. R. Sutherland,1,2 Bernhard J. Hering,1,2 and Zhiguang Guo1,2
1Diabetes Institute for Immunology and Transplantation, University
of Minnesota, Minneapolis, MN
2Department of Surgery, University of Minnesota, Minneapolis, MN
3Klinikum der Albert-Ludwigs-Universität Freiburg, Germany
4Department of Surgery, First Hospital of Beijing University, China
Recently we showed that donor-specific tolerance to MHC-matched islet allografts in diabetic NOD mice could be induced by simultaneous islet and bone marrow transplantation. Mononuclear cell infiltration surrounding the islets was also found in tolerated grafts. In this study, we compared gene expression in the tolerated and rejected islet grafts by using Affymetrix Murine U74A oligonucleotide arrays. To confirm the results of microarray analysis, we performed real-time PCR and RNase protection assay on selected genes. Of over 12,000 genes studied, 57 genes were expressed at consistently higher levels in tolerated islet grafts, and 524 genes in rejected islet grafts. Genes from a variety of functional clusters were found to be different between rejected and tolerated grafts. In the rejected islet grafts, a number of T-cell surface markers and of cytotoxicity-related genes were highly expressed. Also in the rejected grafts, a number of cytokines and chemokines and their receptors were highly expressed. The differential expression of selected genes found by microarray analysis was also confirmed by real-time PCR and RNase protection assay. Our results indicated that gene microarray analysis can help us to detect gene expression differences representative of the biologic mechanisms of tolerance and rejection.
Key words: Gene expression; Islets; NOD mice; Rejection; Tolerance; Graft-infiltrating lymphocytes
Address correspondence to Zhiguang Guo, Islet Transplant Program, University of Minnesota Medical School, 420 Delaware Street SE, Minneapolis, MN 55455. Tel: (612) 626-4571; Fax: (612) 626-5855; E-mail: email@example.com
Individual Human Serum Differs in the Amount of Antibodies With Affinity for Pig Fetal Ventral Mesencephalic Cells and the Ability to Lyse These Cells by Complement Activation
Jan Koopmans,1 Aalzen de Haan,2 Elinda Bruin,2 Ieneke van der Gun,2 Henk van Dijk,3 Jan Rozing,4 Lou de Leij,2 and Michiel Staal1
1Department of Neurosurgery, University Hospital Groningen,
Hanzeplein 1, 9700 RB Groningen, The Netherlands
2Medical Biology Section of Pathology and Laboratory Medicine, University of Groningen, Hanzeplein 1, 9700 RB Groningen, The Netherlands
3Department of Veterinary Anatomy and Physiology, University of Utrecht, Yalelaan 1, 3584 CL, The Netherlands
4Department of Cell Biology, Section Immunology, University of Groningen, A. Deusinglaan 1, 9713 AV Groningen, The Netherlands
Xenografting pig fetal ventral mesencephalic (pfVM) cells to repair the dopamine deficit in patients with Parkinson's disease is the focus of both experimental and clinical investigations. Although there have been marked advances in the experimental and even clinical application of these xenogeneic transplantations, questions regarding the host's xenospecific immune response remain unanswered. It has been shown that human serum is able to lyse pfVM tissue by both anti-gal-gal and non-anti-gal-gal antibodies by complement activation. The aim of this study was to investigate whether interindividual differences exist in the levels of pfVM cell-specific IgM and IgG subclass antibodies, their ability to lyse pfVM cells in vitro and the relationship between both. Pig fetal VM cells were incubated with heat-inactivated serum from 10 different individuals and binding of IgM antibodies and IgG subclass antibodies to pfVM cells was analyzed by flow cytometry. The ability to lyse pfVM cells was analyzed exposing 51Cr-labeled pfVM cells to fresh serum or isolated IgM and IgG from the same individuals and subsequent determination of released 51Cr from lysed cells. Strong differences were found between individuals in the levels of pfVM cell-specific IgM antibodies: antibody levels differed up to 40-fold. pfVM-specific IgG1 and IgG2 levels were only detectable in a few individuals. The ability to lyse pfVM cells ranged from negligible lysis up to 66.5% specific lysis. There was a strong correlation between the levels of individual pfVM-specific IgM antibodies and the ability to lyse pfVM cells in vitro. Isolated IgM, but not IgG, was able to lyse pfVM cells in the presence of complement. In conclusion, the interindividual differences in the levels of IgM with affinity for pfVM cells and their ability to lyse pfVM cells in vitro are considerable. Only few individuals possessed IgG1 and IgG2 subclass antibodies with affinity for pfVM. These findings may influence patient selection for porcine transplants and chances of graft survival in individual patients.
Key words: Xenotransplantation; Ventral mesencephalon; Parkinson's disease; Complement-dependent cytotoxicity; Porcine
Address correspondence to Jan Koopmans, Department of Neurosurgery, University Hospital Groningen, Hanzeplein 1, 9700 RB Groningen, The Netherlands. Tel: +31-50-3613218; Fax: +31-50-3611715; E-mail: J.Koopmans@nchir.azg.nl
Comparison of Intramyocardial and Intravenous Routes of Delivering Bone Marrow Cells for the Treatment of Ischemic Heart Disease: An Experimental Study
Masanori Hayashi, Tao-Sheng Li, Hiroshi Ito, Akihito Mikamo, and Kimikazu Hamano
Division of Cardiovascular Surgery, Department of Medical Bioregulation, Yamaguchi University School of Medicine, Minami-Kogushi 1-1-1, Ube, Yamaguchi, Japan 755-8505
The implantation of bone marrow cells (BMCs) into ischemic heart after myocardial infarction can induce angiogenesis and improve heart function. We compared the advantages of delivering BMCs intramyocardially and intravenously. An acute myocardial infarction model was created by the ligation of left anterior descending artery in female Dark Agouti rats. The rats were then randomly divided into four treatment groups: one given an intramyocardial injection of phosphate-buffered saline (PBS group), one given an intravenous injection of 2 x 107 BMCs from male rats (IV group), one given an intramyocardial injection with total of 2 x 107 BMCs from male rats at four points in the infarction area (IM group), and one given an intravenous injection of 10-fold the number of BMCs from male rats (10xIV group). Quantitative analysis of the SRY gene by real-time PCR showed that the survival of BMCs in the infarcted area was significantly higher in the IM group than in the IV and 10xIV groups, 3 days after treatment (p < 0.05), but not thereafter. However, the blood flow in the infarcted myocardium was significantly better in the IM and 10xIV groups than in the PBS and IV groups 14 days after treatment (p < 0.05). Echocardiography showed that the LVEF continued to decrease in the PBS and IV groups, but was stable after 3 days in the IM and 10xIV groups. By 14 days after treatment, the LVEF was significantly higher in the IM and 10xIV groups than in the PBS and IV groups (p < 0.01). Our results showed that BMCs were more effective delivered intramyocardially than intravenously for inducing angiogenesis and repairing injured myocardium.
Key words: Angiogenesis; Bone marrow cell; Ischemic heart disease
Address correspondence to Kimikazu Hamano, Division of Cardiovascular Surgery, Department of Medical Bioregulation, Yamaguchi University School of Medicine, Minami-Kogushi, Ube, Yamaguchi, Japan 755-8505. Tel: 836-22-2261; Fax: 836-22-2260; E-mail: firstname.lastname@example.org
Viability of Allogeneic Bone Marrow Stromal Cells Following Local Delivery Into Patella Tendon in Rabbit Model
Hong Wei Ouyang,1 James C. H. Goh,1,2 and Eng Hin Lee1
1Department of Orthopaedic Surgery and 2Division of Bioengineering, National University of Singapore, Singapore 119260
Bone marrow stromal cells are potentially useful for tendon repair and regeneration. To provide lasting benefits, the seeded cells must survive implantation at local tendon sites. Our objective was to determine the in vivo fate of allogeneic bone marrow stromal cells (bMSCs) at different time points after implantation into patella tendon defects (i.e., at 2, 3, 5, and 8 weeks). The protocol involved the labeling of bMSCs with green fluorescent protein (GFP) or carboxyfluorescein diacetate (CFDA) before implantation. A window defect (5 x 5 mm) was created at the central portion of rabbit patella tendon and subsequently treated with GFP- or CFDA-marked bMSCs. The marked bMSCs were loaded into the window defect with fibrin glue. Upon sacrifice of the rabbits at the different time points, the implant site of the patellar tendon was immediately retrieved and the viability of the labeled cells was assessed under confocal microscopy. The results showed that the seeded bMSCs remained viable within the tendon wound site for at least 8 weeks after implantation. The cell morphology was changed from a round shape at 2 weeks to a spindle shape at 5 weeks after implantation. This study demonstrated that the bMSCs remained viable for prolonged periods after implantation and therefore have the potential to influence the formation and remodeling of neotendon tissue after tendon repair.
Key words: Viability; Allogeneic; Bone marrow stromal cells; Local delivery; Patellar tendon
Address correspondence to Associate Professor James C. H. Goh, Department of Orthopaedic Surgery, National University of Singapore, 10 Lower Kent Ridge Road, Singapore 119260. Fax: (65)-67744082; E-mail: email@example.com
Reevaluation of Bone Marrow-Derived Cells as a Source for Hepatocyte Regeneration
Tobias Cantz,1* Amar Deep Sharma,1* Andrea Jochheim-Richter,1 Lubomir Arseniev,2** Christoph Klein,3 Michael P. Manns,1 and Michael Ott1
1Laboratory of Cell and Gene Therapy, Department of Gastroenterology, Hepatology and Endocrinology, 2Department of Hematology and Oncology, 3Department of Pediatric Hematology, Hannover Medical School, Hannover, Germany
We have investigated the contribution of intrasplenic bone marrow transplants or in vivo mobilized hematopoietic stem cells to the formation of hepatocytes in normal and injured liver. Direct intrasplenic injections of bone marrow mononuclear cells (5 x 105 cells), Sca1+/lin- hematopoietic stem cells (5 x 103) cells, and highly purified "side population" hematopoietic stem cells (5 x 103) derived from enhanced green fluorescent protein (EGFP)-transgenic mice [C57Bl/6-TgN(ActbEGFP)1Osb] were performed in normal C57Bl/6 mice (n = 6) and in C57Bl/6 mice following two thirds hepatectomy (n = 8). To test the effect of mobilized stem cells on transdifferentiation, C57Bl/6 mice (n = 12) were lethally irradiated and reconstituted with EGFP-positive bone marrow mononuclear cells in a second series of experiments. Eight to 12 weeks after bone marrow transplantation a subset of mice (n = 3 in each group) received either rhG-CSF for hematopoietic stem cell mobilization, rhG-CSF combined with an intraperitoneal application of carbon tetrachloride (CCl4) as hepatocyte regeneration stimulus, or CCl4 alone. All mice were completely perfused with PBS to remove circulating nonorgan cells for analyses 4 weeks later. Liver as well as heart, intestine, spleen, and kidney tissue was analyzed for the presence of EGFP-transgenic cells. In 100 sections (2.3 x 107 cells) of any recipient mouse no EGFP-positive hepatocytes were detected either by analysis of native EGFP fluorescence or by immunofluorescence analysis with anti-EGFP and antidipeptidyl peptidase (DPP) IV antibodies. EGFP-transgenic cells resembling heart, kidney, or intestinal cells could also not be proven. The results demonstrate that there is little or no contribution of bone marrow-derived cells to the regeneration of normal and injured liver in the animal models used. Thus, potential therapeutic prospects of hematopoietic stem cell therapy for liver disease have to be critically reassessed.
Key words: Stem cells; Transdifferentiation; Mobilization; CCl4-induced liver injury; Hepatocyte
Address correspondence to Michael Ott, M.D., Department of Gastroenterology, Hepatology and Endocrinology, Medical School Hannover, 30623 Hannover, Germany. Tel: +49-511-9063544; Fax: +49-511-9063534; E-mail: Ott-MHH@gmx.de
*These authors contributed equally to this work
**Present address: Cytonet Hannover GmbH, Center of Internal Medicine, Hannover, Germany.
Hypothermia Inhibits Fas-Mediated Apoptosis of Primary Mouse Hepatocytes in Culture
Tao Fu,1,3 Andres T. Blei,2 Noriaki Takamura,2 Tesu Lin,2 Danqing Guo,1 Honglin Li,1 Maurice R. O'Gorman,1 and Humberto E. Soriano1
Departments of 1Pediatrics, 2Medicine, and 3Pathology, Northwestern University Feinberg School of Medicine and Children's Memorial Institute for Education and Research, Chicago, IL 60611
Apoptosis occurs during the isolation and even short-term storage and culture of hepatocytes, and in the pathogenesis of liver diseases, such as hepatic failure and hepatitis. Therapeutic hypothermia has beneficial effects in experimental models of fulminant hepatic failure. The mechanisms underlying the potential benefits of mild hypothermia on the liver have not been well investigated. We examined the effects of temperature on soluble Fas ligand-induced apoptosis in freshly isolated mouse hepatocytes. Decreasing the culture temperature from 37°C to 32°C produced significant suppression of Fas-mediated apoptosis in cultured hepatocytes over a 12-h period. This observation was supported by cell morphology, flow cytometry analysis of cellular DNA content, and Annexin V-FITC staining of membrane phosphatidylserine translocation. In hypothermic conditions, Fas-mediated cytochrome c release from mitochondria of hepatocytes and the proximate downstream activation of caspase-9 were suppressed under mild hypothermic conditions. Effector caspase-7 activity was also inhibited at 32°C. In contrast, the activation of initiator caspase-8 and cleavage of Bid were not affected after Fas-ligand stimulation. These findings suggest that mild hypothermia suppresses Fas-mediated apoptosis of liver cells by the partial inhibition of signaling events including mitochondrial damage, cytochrome c release, and subsequent apoptosome formation and effector caspase activation.
Key words: Hypothermia; Hepatocyte; Apoptosis; Fas protein; Caspases
Address correspondence to Tao Fu, M.D., Ph.D., Department of Pathology, Northwestern University, Feinberg School of Medicine, 303 East Chicago Avenue, Chicago, IL 60611. Tel: (312) 503-1440; Fax: (312) 503-8240; E-mail: firstname.lastname@example.org
Optimization of Conditions for Clinical Human Hepatocyte Infusion
Robert A. Fisher,1 Dawen Bu,1 Melissa Thompson,1 Luke Wolfe,1 and Joseph K. Ritter2
Department of 1Surgery and 2Pharmacology & Toxicology, Virginia Commonwealth University Medical Center, Richmond, VA 23298
Cytotoxicity and apoptosis are common problems in the isolation and storage of human hepatocytes. In vitro environments of hepatocytes during cell infusion may be critical to reducing cellular damage and enhancing cell viability. We examined the effects of donor liver histology (40-50% steatosis vs. normal), incubation time, temperature, and three solutions for infusion on banked primary human hepatocytes, by studying: trypan blue exclusion, AST release, LDH release, MTT assay, detection of DNA ladder, and a hepatocyte proliferation assay. In addition, the microstructure functions of the endoplasmic reticulum and mitochondria of the intact hepatocytes were determined by measuring correlates of UGT 1A1 and cytochrome P-450 3A (CYP3A4) activity. In general, hepatocyte viability decreased significantly within 60 min after thawing. Cells suspended in 5% dextrose lactated Ringers solution (D5LR) maintained greater cell viability. Hepatocytes from normal liver donors showed less AST and LDH enzyme leak in comparison with cells from fatty liver donors. Mild hypothermic temperature (32°C) inhibited cellular damage that otherwise significantly increased at 60 min. Hepatocytes did not proliferate until 12 h from thaw, regardless of supernatant or conditions of suspension. CYP3A4 activity and a marker for UGT 1A1 activity in hepatocytes from normal donor livers were higher than those from steatotic donor livers. These findings suggest that hepatocytes suspended for infusion after isolation from normal liver donors have normal biological functions and less cellular damage/necrosis in contrast with those isolated from fatty liver donors. These damages are inhibited significantly by maintaining hepatocytes at a mild hypothermic temperature (32°C). D5LR alone maintained the best cell viability for up to 60 min. Media of D5LR + adenosine and HMM were able to partially inhibit hepatocyte apoptosis in hepatocytes from steatotic livers.
Key words: Primary human hepatocytes; Hypothermia; Cytotoxicity; Apoptosis; Transplantation; Hepatocyte transplantation; Fatty liver donor; Culture media of hepatocytes
Address correspondence to Robert A. Fisher, M.D., Division of Transplantation, Department of Surgery, P.O. Box 980254, Virginia Commonwealth University Medical Center, Richmond, VA 23298. Tel: (804) 828-2461; Fax: (804) 828-2462; E-mail: email@example.com
Transplantation of Cultured Salivary Gland Cells Into an Atrophic Salivary Gland
T. Sugito,1 H. Kagami,2 K. Hata,3 H. Nishiguchi,1 and M. Ueda1
1Department of Oral and Maxillofacial Surgery, Nagoya University
Graduate School of Medicine, Nagoya 466-8440, Japan
2Department of Tissue Engineering, Nagoya University School of Medicine, Nagoya 466-8550, Japan
3Center for Genetic and Regenerative Medicine, Nagoya University Hospital, Nagoya 466-8550, Japan
Patients with dry mouth have been treated with salivary substitutes and/or medications such as pilocarpine or cevimeline hydrochloride. These treatments temporarily relieve their symptoms and induce salivation from residual tissue. However, no treatment is available for the purpose of regenerating an atrophic gland. In this study, the feasibility of a cell transplantation therapy for the atrophic submandibular glands was investigated in rats. Further, the potential of cell differentiation into a useful phenotype was assessed by immunohistochemistry together with cell tracking with the fluorescent dye PKH 26. Rat submandibular glands were excised, and the salivary gland epithelial cells were cultured for 3 weeks with 3T3 cells as a feeder layer. Ductal ligation of the submandibular gland was employed to generate an atrophic gland. One week after the operation, the ligation was removed, and the cultured cells labeled with PKH 26 were injected into the atrophic submandibular glands. As a control, the cultured cells were also injected into normal submandibular glands. Two weeks after cell transplantation, the transplanted cells were detectable in both the experimental and control groups. The cells were clustered in the connective tissue between the lobules. Four weeks after transplantation, the labeled cells were detectable in the experimental group but not in the control group. In the atrophic glands, the scattered transplanted cells were observed over a broad area of the gland but localized mainly around the acini and ductal region. Immunostaining results showed a possible involvement of the transplanted cells in ductal regeneration, while neither myoepithelial nor acinar differentiations were observed within the 4 weeks since transplantation. This study demonstrated that cell transplantation to the salivary gland is feasible, and that the transplanted cells were selectively attracted to and remained in the damaged area without affecting normal tissue.
Key words: Cell transplantation; Tissue engineering; Salivary gland; Cultured salivary gland cells; Atrophic gland
Address correspondence to H. Kagami, Department of Tissue Engineering, Nagoya University School of Medicine, 65 Tsuruma-cho, Showa-ku, Nagoya 466-8550, Japan. Tel: 052-744-2424; Fax: 052-744-1978; E-mail: firstname.lastname@example.org
Allogeneic Versus Xenogeneic Immune Reaction to Bioengineered Skin Grafts
Gulsun Erdag1 and Jeffrey R. Morgan1,2
1Center for Engineering in Medicine and Surgery, Massachusetts
General Hospital, Harvard Medical School and Shriners Hospital for Children,
Boston, MA, 02114
2Brown University, Providence, RI, 02912
There are conflicting reports on the survival and immune reaction to allografts and xenografts of cultured skin substitutes (CSS). In this study, we investigated the allogeneic and xenogeneic responses to CSS of human keratinocytes and genetically engineered CSS expressing keratinocyte growth factor (KGF) that forms a hyperproliferative epidermis. CSS (control and KGF modified) and neonatal human foreskins were evaluated by immunohistochemistry for the expression of MHC class I and II. To study allograft rejection, grafts were transplanted to human peripheral blood mononuclear cell (huPBMC)-reconstituted SCID mice. To study xenograft rejection, grafts were transplanted to immunocompetent mice. Graft survival and immune reaction were assessed visually and microscopically. After transplantation, control CSS formed a normal differentiated epidermis, whereas KGF CSS formed a hyperproliferative epidermis. Control and KGF CSS expressed class I similar to neonatal foreskin, but did not express class II. In the allograft model, rejection of neonatal foreskins was between 5 and 9 days. In contrast, neither control nor KGF CSS was rejected by huPBMC-SCID mice. Histology showed dense mononuclear cell infiltration in human foreskins, with few, if any, mononuclear cells in control or KGF CSS. In contrast to the allogeneic reaction, CSS (control and KGF) were rejected in the xenograft model, but rejection was delayed (9-21 days) compared with neonatal skin (5-8 days). Humanized SCID mice rejected allografts of human neonatal foreskins, but did not reject control CSS or KGF CSS, even though the KGF CSS formed a hyperproliferative epidermis. Rejection of control and KGF CSS by immunocompetent mice in a xenograft model was comparable and their survival was significantly prolonged compared with neonatal skin. These results demonstrate that control CSS and hyperproliferative KGF CSS are less immunogenic than normal human skin and that sustained hyperproliferation of the epidermis does not accelerate rejection.
Key words: Cultured skin substitute; Rejection; Allograft; Xenograft; Keratinocyte growth factor
Address correspondence to Jeffrey R. Morgan, Ph.D., Brown University, G-B 393, Biomed Center, 171 Meeting St., Providence, RI 02912. Tel: (401) 863-9879; Fax: (401) 863-1753; E-mail: Jeffrey_Morgan@Brown.edu
Improved Histological Evaluation of Vascularity Around an Immunoisolation Device by Correlating Number of Vascular Profiles to Glucose Exchange
Anne Sörenby,1 Ehab Rafael,1 Annika Tibell,1 and Annika Wernerson2
Departments of 1Transplantation and 2Pathology, Karolinska Institutet, Karolinska University Hospital Huddinge, Stockholm, Sweden
The aim of this study was to determine which vessels are important for the exchange of small molecules, such as glucose, from the microcirculation into an immunoisolation device. Reasonably, those vessels should be the ones of interest in histological evaluations. In a previous study, we examined the diffusion of glucose from the microcirculation into immunoisolation devices that had been implanted subcutaneously in rats for various times (i.e., 1, 2, and 4 weeks and 3 months). The glucose kinetic data were then correlated with the number of vascular profiles within 15 and 250 mm from the device. Significant correlations were found only at 250 mm. To examine the relation further between function and vascularization, we used the histological samples from the previous study and counted vascular profiles within various distances between 15 and 400 mm from the device. The number was then correlated with the already available glucose kinetic data. The highest correlations were found at 75 and 100 mm (p < 0.05). We therefore suggest that vascular profiles within 100 mm should be used when evaluating the vascularity of tissue surrounding an immunoisolation device. We also studied neovascularization asymmetries between the side of the membrane facing the skin and that facing the muscle. At 1 and 2 weeks about half of the devices were mainly vascularized on the side facing the skin, whereas the rest were equally vascularized on the two sides. At 3 months, all devices were well vascularized, and no striking vascularization asymmetries were seen.
Key words: Biocompatibility; Cell transplantation; Immunoisolation; Microcirculation; Neovascularization
Address correspondence to Anne Sörenby, M.D., Department of Transplantation Surgery, Karolinska Institutet, Karolinska University Hospital Huddinge, 141 86 Stockholm, Sweden. Tel: +46 734 22 89 09; Fax: +46 8 774 31 91; E-mail: email@example.com