Alloimmunity to Human Endothelial Cells Derived from Cord Blood Progenitors

This information is current as of January 15, 2019. Alloimmunity to Human Endothelial Cells Derived from Cord Blood Progenitors Yajaira Suárez, Benjamin R. Shepherd, Deepak A. Rao and Jordan S. Pober J Immunol 2007; 179:7488-7496; ; doi: 10.4049/jimmunol.179.11.7488 http://www.jimmunol.org/content/179/11/7488 References Subscription Permissions Email Alerts This article cites 45 articles, 31 of which you can access for free at: http://www.jimmunol.org/content/179/11/7488.full#ref-list-1 Why The JI? Submit online. Rapid Reviews! 30 days* from submission to initial decision No Triage! Every submission reviewed by practicing scientists Fast Publication! 4 weeks from acceptance to publication *average Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Submit copyright permission requests at: http://www.aai.org/about/publications/ji/copyright.html Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts Downloaded from http://www.jimmunol.org/ by guest on January 15, 2019 The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright 2007 by The American Association of Immunologists All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606.

The Journal of Immunology Alloimmunity to Human Endothelial Cells Derived from Cord Blood Progenitors 1 Yajaira Suárez, Benjamin R. Shepherd, Deepak A. Rao, and Jordan S. Pober 2 There is considerable interest in exploiting circulating endothelial progenitor cells (EPCs) for therapeutic organ repair. Such cells may be differentiated into endothelial cells (ECs) in vitro and then expanded for use in tissue engineering. Vessel-derived ECs are variably immunogenic, depending upon tissue source, and it is unknown whether ECs derived from cord blood EPCs are able to initiate an allogeneic response. In this study, we compare the phenotype and alloantigenicity of human cord blood progenitor cell-derived ECs with HUVECs isolated from the same donors. Human cord blood progenitor cell-derived ECs are very similar to HUVECs in the expression of proteins relevant for alloimmunity, including MHC molecules, costimulators, adhesion molecules, cytokines, chemokines, and IDO, and in their ability to initiate allogeneic CD4 and CD8 memory T cell responses in vitro and in vivo. These findings have significant implications for the use of cord blood EPCs in regenerative medicine or tissue engineering. The Journal of Immunology, 2007, 179: 7488 7496. Both the regeneration of injured tissues and the construction, through bioengineering, of synthetic tissue substitutes depend upon the formation of new blood vessels (1 4). Endothelial cells (ECs) 3 isolated from the luminal lining of mature blood vessels can spontaneously organize into tubes that may serve as the nucleus for new vessel development (3 6). However, vessel-derived ECs have only limited replication potential and rapidly senesce when expanded in vitro (7). ECs may also be differentiated from endothelial progenitors cells (EPCs) that are found in peripheral blood (8). Umbilical cord blood is a particularly rich source of EPCs that can be used for tissue engineering (9, 10). The EPCs found in human cord blood may themselves be heterogeneous. Some EPCs, when cultured in basal medium supplemented with FCS and vascular endothelial growth factor (VEGF), rapidly produce colonies of EC-like cells. These early outgrowth cells express some EC markers, like CD31 or von Willebrand factor, but also express markers of monocytes, like CD14 and CD45. Such cells have limited replicative capacities and do not form stable tubes when implanted in immunodeficient mice (11 13,15). Other EPCs give rise to EC-like colonies only after a delay of 7 days or more (10, 14, 16 21). Such late outgrowth cells have high replicative capacity, express EC but not monocyte markers, and form stable tubes in vivo (14). Hereafter we will refer to the differentiated progeny of late outgrowth EPCs as human cord Department of Immunobiology, Interdepartmental Program in Vascular Biology and Therapeutics, Yale University School of Medicine, New Haven, CT 06520 Received for publication June 26, 2007. Accepted for publication September 21, 2007. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 1 This work was supported in part by National Institute of Health Grants HL-51015, HL85416, and HL-70295 and Programme 3 3 Fellowship from the Centro Nacional de Investigaciones Cardiovasculares (CNIC), Spain (to Y.S.). 2 Address correspondence and reprint requests to Dr. Jordan S. Pober, Yale University School of Medicine, 10 Amistad Street, Room 401D, New Haven, CT 60509. E-mail address: jordan.pober@yale.edu 3 Abbreviations used in this paper: EC, endothelial cell; EPC, endothelial progenitor cell; HUVEC, human umbilical veins; HCBEC, human cord blood progenitor cellderived EC; VEGF, vascular endothelial growth factor; Uea-1, Ulex europeus agglutinin 1; MFI, mean fluorescence intensity. Copyright 2007 by The American Association of Immunologists, Inc. 0022-1767/07/$2.00 blood derived ECs (HCBECs). Because of their greater replicative potential and their ability to form stable tubes in vivo, late outgrowth cells are a much more useful source of ECs for therapeutic purposes than early outgrowth EPCs. Immunological rejection of allogeneic stem cells following differentiation represents a major hurdle for both organ regeneration and tissue engineering. This is likely to be particularly true for vascularized structures because some vessel-derived human ECs are capable of initiating an allogeneic immune response, directly presenting both non-self allelic forms of class I and class II MHC molecules to alloreactive memory CD8 and CD4 T cells, respectively (22, 23). Specifically, coculture of purified peripheral blood memory T cells subsets with allogeneic vessel-derived human ECs leads to expression of T cell activation markers, cytokine production, and proliferation (21, 22, 24, 25). Moreover, experiments using immunodeficient mouse hosts have revealed that human ECs are capable of triggering graft rejection by adoptively transferred alloreactive effectors or memory T cells (24). The majority of experiments involving vessel-derived ECs have used either cells isolated from human umbilical veins (HUVEC) or human dermal microvessels. It has been reported that ECs from other vascular beds may differ from HUVECs in their capacities to activate allogeneic T cells (26). It is unknown whether ECs derived from cord blood EPCs are able to activate allogeneic T cells, and if they can do so, how they compare with other types of ECs isolated from mature vessels. In the present study, we have compared HCBECs with HUVECs in a variety of assays for immunological functions. In most instances, we have been able to isolate HCBECs and HUVECs from the same umbilical cord to control for genetic variations. Our key finding is that HCBECs are largely indistinguishable from HUVECs in their expression of molecules of relevance to allogeneic T cell activation and in their ability to stimulate an alloresponse, a result that has important implications for the use of HC- BECs for regenerative medicine or tissue engineering Materials and Methods Cell culture All human cell populations were obtained using protocols approved by the Yale Human Investigation Committee. HUVECs were released from cannulated and perfusion-cleared umbilical veins, by collagenase digestion, www.jimmunol.org

The Journal of Immunology 7489 and serially cultured on 0.1% gelatin-coated flasks in M199/20% FBS supplemented with L-glutamine, penicillin/streptomycin (Invitrogen Life Technologies), and endothelial cell growth supplement (Calbiochem) with heparin from porcine intestine as described (27). HCBECs were differentiated from cord blood mononuclear cells in vitro as late outgrowth cells, as previously described (10), and then expanded in culture. Cultures were serially propagated on gelatin-coated flasks in EGM-2/15% FBS and stepwise transitioned to grow in HUVEC culture medium. Cell surface immunostaining ECs were analyzed for surface expression of specific markers by fluorescence flow cytometry. Where indicated, cultured EC were exposed to 10 ng/ml TNF- (R&D Systems) or 50 ng/ml IFN- (BioSource International) for the times specified in the text before harvest. Confluent monolayers were washed twice in HBSS (Invitrogen Life Technologies) and then incubated with trypsin-edta for 1 min. Enzyme activity was then quenched with 20% FBS/M199 and suspended cells were collected and washed in 10 ml of cold 1% BSA (Sigma-Aldrich)/PBS, centrifuged, and washed again in cold 1% BSA/PBS. Cells were then incubated with 2 g/ml primary Ab or isotype control directly conjugated to FITC or PE for 45 min. Immunostained cells were then washed twice with cold PBS and analyzed on a FACSort flow cytometer (BD Biosciences) using CellQuest analysis software collecting 10,000 gated viable cells per sample. Specific Abs used in these analyses were reactive with human CD31, CD45, CD14, HLA-ABC, HLA-DR, CD40, CD80, and CD86 from Beckman Coulter, with CD34 from Miltenyi Biotec, with VEGFR2, glucocorticoid-induced TNF receptor ligand, and E-selectin from R&D Systems, and with CD58 (LFA-3), PD-L1 (CD274), PD-L2 (CD273), 41BB ligand (CD137L), Ox-40 ligand (CD134L), VCAM-1, and ICAM-1 from BD Pharmingen. Anti-human ICOS ligand mab PE-labeled was a gift from H. W. Mages (Forschungs Institut for Molekulare Pharmacologie, Berlin, Germany). Alternatively, cells were stained with FITC-conjugated Ulex europeus agglutinin (Uea-1) (Sigma-Aldrich), which reacts with the blood group ABH expressed on human ECs. Western blot analysis HUVECs or HCBECs were lysed in ice-cold buffer (containing 50 mm Tris-HCl (ph 7.5), 10% glycerol, 125 mm NaCl, 1% Nonidet P-40, 5.3 mm NaF, 1.5 mm Na 2 PO 4 ), 1 mm orthovanadate, 175 mg/ml octylglucopyranoside, 1 mg/ml protease inhibitor mixture (Roche), and 0.25 mg/ml 4-(2-aminoethyl)-benzenesulfonyl fluoride (Roche). Cell lysates were rotated at 4 C for 30 min before the insoluble material was removed by centrifugation at 12,000 g for 10 min. After normalizing for equal protein concentration, cell lysates were heated in SDS sample buffer before separation by SDS-PAGE. Following overnight transfer of the proteins onto nitrocellulose membranes, Western blots were performed using the Table I. mabs reactive with human IDO and Hsp-90 (loading control) from Chemicon International and BD Biosciences, respectively. Cytokine and chemokine assays Supernatants of cultured ECs exposed to 10 ng/ml TNF- (R&D Systems), 50 ng/ml IFN- (BioSource International), or mock-treated for the times specified in the text were collected, and the samples were assessed for IP-10 or IL-8 using ELISA kits from BD Biosciences and BioSource International, respectively. Cell lysates of cultured ECs were prepared as indicated in Western blot analysis section and were used to analyze the expression of IL-1 using an ELISA kit from PeproTech. T cell isolation PBMCs were isolated by density gradient centrifugation of leukapheresis products by using Lymphocyte Separation Medium (Invitrogen Life Technologies) according to the directions of the manufacturer. Isolated PBMCs were stored in 10% DMSO-90% FBS at 196 C, thawed, and washed before use (28). CD4 or CD8 T cells were isolated from PBMCs by positive selection using Dynabeads (Dynal Biotech). The selected population obtained by this procedure was routinely 95% positive for the desired marker by flow cytometry (data not shown). Activated T cells and monocytes were removed by negative selection using an anti-hla-dr mab at a concentration of 5 g/ml (LB3.1; gift of J. Strominger, Harvard University, Cambridge, MA) for 20 min, washed twice, and depleted by using magnetic beads conjugated to goat anti-mouse Ab (Dynal Biotech). Naive and memory subsets of CD4 and CD8 T cells were isolated from the purified T cell subset populations by further negative selection using anti-cd45ro or anti-cd45ra mabs, respectively, at a concentration of 2 g/ml (BioSource International). The selected subset population obtained by this procedure was routinely 95% positive for the nondepleted CD45R isoform by flow cytometry (data not shown). T cell-ec cocultures ECs ( 1.5 10 5 cells) were plated into gelatin-coated wells of 24-well culture plates (Falcon; BD Biosciences) and treated with IFN- (50 ng/ml) (BioSource International) where indicated. Purified T cell subsets were then added to each well ( 2 10 6 per well). All cultures were maintained in 5% CO 2 at 37 C. The medium for coculture consisted of RPMI 1640 supplemented with 10% FBS serum, 2 mm L-glutamine, 100 U/ml penicillin, and 100 g/ml streptomycin all from Invitrogen Life Technologies. Supernatants collected from cocultures of T cells and ECs after the indicated times of coculture were assayed using an ELISA kit for IFN- (BioSource International) or IL-2 (ebioscience). All ELISAs were performed as described by the manufacturers. To measure proliferation by CFSE dilution, CD4 or CD8 T cells were stained with 250 nm CFSE (Molecular Probes) for 20 min before coculture Flow cytometric analysis of proteins of immunological significance in HUVECs and HCBECs isolated from the same donor a HUVEC HCBEC Control IFN (36 h) TNF (6 h) TNF (24 h) Control IFN (36 h) TNF (6 h) TNF (24 h) PECAM-1 (CD31) 353 358 377 329 369 362 382 348 CD45 ND ND ND ND ND ND ND ND CD14 ND ND ND ND ND ND ND ND HLA-ABC 54 749 88 247 50 688 74 285 HLA-DR ND 27 ND ND ND 23 ND ND LFA-3 (CD58) 28 29 31 39 29 25 27 30 CD40 35 61 36 38 16 53 19 47 B7-1 (CD80) ND ND ND ND ND ND ND ND B7-2 (CD86) ND ND ND ND ND ND ND ND ICOS ligand 18 23 35 101 22 19 36 100 4-1BB ligand (CD137L) 20 19 20 21 16 19 20 18 Ox40 ligand (CD134L) 337 331 390 343 102 69 128 76 GITR ligand 7 9 11 8 10 13 10 4 PD-L1 (CD274) 11 37 12 15 17 50 22 22 PD-L2 (CD273) 13 54 18 23 19 53 28 49 VCAM-1 (CD106) ND 1 20 39 2 6 21 52 ICAM-1 (CD54) 80 526 971 1174 187 363 695 1478 E-selectin (CD62E) 1 2 48 11 ND 1 28 5 a Cells were exposed to 10 ng/ml TNF, 50 ng/ml IFN, or mock treated for the indicated times. Cells were harvested and labeled as described in Materials and Methods. The results are expressed in terms of corrected median fluorescence intensity (MFI), in arbitrary units of fluorescence, after subtracting the MFI of isotype control from the MFI of the specifically stained cells. Each cell type was analyzed from at least two separate cultures with similar results. ND, Not detected.

7490 ALLOIMMUNITY TO ENDOTHELIUM FROM CORD BLOOD PROGENITORS FIGURE 1. Time course of surface expression of E-selectin, VCAM-1, and ICAM-1 after cytokine treatment on HCBECs and HUVECs. HUVECs (f) and HCBECs ( ) isolated from the same donor were exposed to 10 ng/ml TNF (continuous line) or 50 ng/ml IFN- (dotted line) for the indicated times. Cells were harvested and labeled as described in Materials and Methods. The results are expressed in terms of corrected MFI, in arbitrary units of fluorescence, after subtracting the MFI of isotype control from the MFI of the specifically stained cells. with EC. Following 1 wk of coculture, T cells were collected and stained with anti-cd4 or CD8 PE-labeled mabs (Beckman Coulter), respectively, and analyzed using two-color FACS. Duplicate samples were labeled with anti-cd25 PE-labeled mab, also from Beckman Coulter. In vivo analysis of EC alloantigenicity All animal protocols were approved by the Yale Institutional Animal Care and Use Committee. Human microvessels were generated and implanted in the s.c. position on the abdominal wall of C.B-17/Scid-beige mice as previously described (5). Briefly, HCBECs or HUVECs were suspended in a rat tail type I collagen-human plasma fibronectin gel and 650 l ofthe cell suspension was gently poured into a single well of a 24-well tissue culture plate. The protein gel was polymerized at 37 C and an equal volume of EGM-2MV was added to the well. Twelve to 18 h after gel polymerization, the gels were removed, bisected, and implanted in the s.c. position on the abdominal wall. Three weeks after implantation, one third of the animals were euthanized and the grafts harvested for analysis of the human microvasculature. A second third of animals were given 3 10 8 allogeneic human PBMCs by intraperitoneal inoculation. The remaining animals were injected with PBS and enrolled as nonreconstituted controls. Before the termination of the experiment, the number of circulating human T cells was evaluated. Heparinized retro-orbital venous blood samples were obtained and the erythrocytes were lysed using ammonium chloride. Leukocytes were incubated with PE-conjugated rat anti-mouse CD45 (BD Pharmingen) mab and FITC-conjugated mouse anti-human CD3 mab (Immunotech/Beckman Coulter) in PBS/1% BSA/5% NGS plus 2 g/ml FIGURE 2. Chemokine secretion (IL-8 and IP-10), inflammatory cytokine expression (IL-1 ), and the immunoinhibitory enzyme IDO expression in HUVECs and HCBECs isolated from the same donor. Cells were exposed to 10 ng/ml TNF (u), 50 ng/ml IFN- (F), or mock-treated (E) for 24 h. Supernatants were collected and the samples were assessed for IL-8 (A) or IP-10 (B) using ELISA as described in Materials and Methods. Cells were treated as indicated and cell lysates of cultured ECs were prepared as indicated in Materials and Methods. Lysates analyzed for expression of IL-1 (C) using ELISA and IDO (D) by Western blottign. Hsp-90 protein levels were used as loading control. For panels A C, data represent mean SEM of quadruplicate determinations from one of two independent experiments with similar results. For panel D, blots are from one of two independent experiments with similar results.

The Journal of Immunology 7491 FIGURE 3. Activation of CD4 T cells by allogeneic HUVECs and HCBECs isolated from the same donor (donors I and II). HUVECs and HCBECs were treated with IFN- to up-regulate MHC class I/II or mock-treated and subsequently cocultured with CD4 T cells. A and B, Medium were collected for ELISA from HUVECs and HCBECs cultured with CD4 T cells for 24 h. Significant production of both IFN- (A and B, left panels) and IL-2 (A and B, right panels) was observed when compared with resting ECs (not pretreated with IFN- ) orcd4 T cells cultured in the absence of ECs. Data are mean SEM of quadruplicate determinations in one of two independent experiments with similar results. C, After 1 wk of cocultures, CFSE-labeled T cells were stained with anti-cd4 and anti-cd25 mabs and subjected to FACS analysis. CD4 T cells demonstrate significant proliferation in HUVECs or HCBECs cocultures of Donor I and Donor II, as judged by CFSE dilution, and CD25 expression was enhanced in proliferating cells (upper and lower panels, respectively). Data are from one of two independent experiments with similar results. Fc-Block (BD Biosciences) and analyzed by FACS. Three weeks following PBMC inoculation, all remaining mice were euthanized and the grafts harvested for analysis of human EC-lined microvessels and the degree of allogeneic PBMC infiltration. Recovered gels and surrounding soft tissue were fixed in 10% buffered formalin and embedded in paraffin. Sections (5- m thick) were cut for immunostaining and for H&E staining. In some experiments, the harvested grafts were bisected and half of the specimen was fixed in 10% buffered formalin and half snap frozen in Tissue-Tek OCT (Sakura Finetek), the latter used to prepare 6- m cryosections. Sections were also stained with tetramethylrhodamine isothiocyanate-uea-1- labeled lectin for detection of human EC-lined vessels within engrafted protein gels. Statistical analysis All data are expressed as means SEM. Statistical differences were measured by Student s t test. A value of p 0.05 was considered as statistically significant. Results Phenotypic analyses We first characterized multiple individual isolates of both HC- BECs and HUVECs by extensive phenotyping. HCBECs were differentiated from EPCs in vitro as late outgrowth colonies and then serially expanded in culture (10). As we previously reported (10), these HCBECs formed a confluent monolayer of nonoverlapping polygonal cells containing cytoplasmic granules that stained positive for von Willebrand factor and formed intercellular junctions expressing vascular endothelial-cadherin. By flow cytometry, HCBECs uniformly bound Uea-1 and expressed CD34, VEGFR2 and CD31, all markers of differentiated ECs. The HC- BECs used in these experiments were uniformly negative for myeloid markers CD133, CD45, CD18, and CD14 (data not shown), which may be coexpressed by early outgrowth ECs (14). This pattern of lineage marker expression was indistinguishable from that of HUVECs and defines HCBECs as true endothelium. However, compared with HUVECs, HCBECs grew faster (data not shown) and appeared capable of a greater number of population doublings without evidence of senescence, consistent with the growth characteristics previously reported for late outgrowth EPCderived ECs (14, 16, 21). Because of their robust growth potential, HCBECs are of special interest to tissue engineering. Having established that HCBECs appear to be true ECs, we proceeded to analyze the expression of surface proteins of immunological significance under basal conditions and after cytokine (TNF or IFN- ) activation. There is considerable genetic variability among humans in EC expression of molecules of immunological relevance (29). Therefore, six independent isolates were compared and the results were generally consistent with some minor quantitative variations. This extent of variability among donors was comparable to the variability among HUVECs from different donors (data not shown). To compare HCBECs with HUVECs, we isolated three pairs from the same donor, avoiding donor-dependent variability. One representative data set from one of these experiments is shown in Table I. The expression levels in HCBECs of the strongest stimulators of alloimmune response, class I and class II MHC molecules (HLA-ABC and HLA-DR, respectively), were very similar to those in HUVECs, showing only low levels of class I MHC molecules in resting cells and expressing more class I and de novo induction of class II MHC molecules after IFN- treatment (30). Additionally, TNF induced-increases of class I MHC molecules (31) were very similar in both EC types. HCBECs expressed the same set of costimulatory molecules as HUVECs, namely CD58, CD40, ICOS ligand, 4-1BB ligand, Ox40 ligand,

7492 ALLOIMMUNITY TO ENDOTHELIUM FROM CORD BLOOD PROGENITORS FIGURE 4. Activation of CD8 T cells by allogeneic HUVECs and HCBECs isolated from the same donor. HUVECs and HCBECs were cocultured with CD8 T cells. A and B, Medium were collected for ELISA from HUVECs and HCBECs cultured with CD8 T cells for 48 h. CD8 T cells cultured with both HUVECs and HCBECs showed significant production of both IFN- (A) and IL-2 (B) when compared with CD8 T cells cultured in the absence of ECs. Data are mean SEM of quadruplicate determinations from one of three independent experiments with similar results. C, After 1 wk of cocultures, CFSE-labeled T cells were stained with anti-cd8 or anti-cd25 mabs and subjected to FACS analysis. CD8 T cells demonstrate significant proliferation in HUVECs and HCBECs cocultures, as judged by CFSE dilution (4.3% and 3.5%, respectively), and the CD25 expression was enhanced in proliferating cells (upper and lower panels, respectively). Data are from one of three independent experiments with similar results. and glucocorticoid-induced TNF receptor ligand, but not CD80 or CD86 (Table I). The levels of expression were generally comparable between HCBECs and HUVECs. The single notable difference being that basal expression of CD40 was consistently lower on HCBECs than on HUVECs, although comparable expression levels were seen after cytokine activation (32). The expression of the negative signaling molecules PD-L1 and PD-L2 on HCBECs were also very similar to HUVECs. We next compared the expression levels of the TNF-induced leukocyte adhesion molecules E-selectin, VCAM-1, and ICAM-1. No significant differences were observed between HCBECs and HUVECs. The pattern and level of expression of these adhesion molecules following either TNF or IFN- treatment was analyzed in a time course experiment and found to be similar between both cell types (Fig. 1). Cumulatively, these data suggest that HCBECs express the same pattern of surface molecules relevant for interactions with the immune system as HUVECs. Many soluble factors secreted by ECs may also contribute to allogeneic responses. Therefore, we analyzed the EC secretion of some of these proteins including cytokine-induced proinflammatory chemokines (IL-8 and IP-10) and the inflammatory cytokine (IL-1 ) (in our hands HUVECS synthesize very little IL- and do not synthesize TNF (33) and unpublished data). As shown in Fig. 2, A and B, the secretion of IL-8 and IP-10 in response to TNF and IFN-, respectively, was very similar in both cell types and among different donors. The levels of IL-1 after TNF treatment were also very similar in both cell types but varied more highly among donors (Fig. 2C). We also examined the expression of the immunoinhibitory enzyme IDO, which may play a role in the regulation of adaptive immune responses (26, 34). The level of expression of IDO was undetectable under basal conditions but was induced to comparable levels in both cell types following IFN- treatment (Fig. 2D). Additionally, when the surface expression of CD95L (FasL) was analyzed under basal conditions and after cytokine treatment, no detectable expression was found in either cell type (data not shown). Functional analyses in vitro Cultured HUVECs can present both class I and class II MHC molecules in a manner that results in the activation of allogeneic memory T cells (25, 35). We analyzed the responses of multiple T cell isolates from adult human peripheral blood to multiple different cultures of HCBECs and HUVECs and found generally similar responses to both cell types (data not shown). Once again, to avoid differences attributable to different donor origin we focused our experiments on the response of T cells to pairs of HCBECs and HUVECs isolated from the same EC donors, and the results from two independent donors are shown in Fig. 3. In these experiments with CD4 T cells, HUVECs and HCBECs were first treated with IFN- for three days (to induce HLA-DR expression) or mocktreated and then cocultured with allogeneic, CFSE-labeled CD4 T cells. Untreated (HLA-DR ) HUVECs and HCBECs from both donor I and II were less able to induce CD4 T cells to secrete either IFN- or IL-2 (Fig. 3, A and B). In contrast, both types of ECs that had been pretreated with IFN- induced significant production of these two cytokines in cocultures with allogeneic CD4 T cells (Fig. 3, A and B). The overall degree of secretion of both cytokines by CD4 T cells was higher in response to donor I than in donor II. Although cytokine secretion by T cells stimulated with EC from donor I was slightly less in HCBEC cocultures than in HUVEC cocultures (Fig. 3A, left and right panels), this difference was not observed in cocultures donor II (Fig. 3B, left and right panels). Consistent with IL-2 secretion, CD4 T cells demonstrate both proliferation and activation in response to IFN- -pretreated HUVECs or HCBECs from both donor I and II, as judged by CFSE dilution and CD25 expression in proliferating cells, respectively (Fig. 3C, upper and lower panels). Again, the degree of the T cell response was higher in cocultures with ECs from donor I

The Journal of Immunology and HCBEC cocultures from this donor were less able to activate CD4 T cells than HUVEC cocultures from this same donor. This difference was less clear in ECs from donor II. These results show that HCBECs are roughly comparable to HUVECs in their ability to activate allogeneic CD4 T cells. We next compared the capacity of HCBECs to activate allogeneic CD8 T cells. As shown in Fig. 3, resting HCBECs were able to induce CD8 T cells to secrete IFN- and IL-2 (Fig. 4, A and B) and to proliferate (Fig. 4C). As expected, similar percentages of proliferating cells were found when activated cells were analyzed (CD25 high and CFSE low ) (Fig. 4C, lower panels). The CD8 T cell response to HCBECs was similar to the response to HUVECs in the same experiments. HUVECs preferentially activated allogeneic memory T cells compared with naive T cells (22, 23). To assess whether HCBECs 7493 also share this characteristic, allogeneic CD4 and CD8 T cells were first separated into memory (CD45RO ) and naive (CD45RA ) T cells and then separately cocultured with HCBECs or HUVECs (Fig. 5). As shown in Fig. 4, A and B, CD4 T cells (memory and naive) did not respond to HUVECs or HCBECs that were not pretreated with IFN-, but IFN- pretreated HUVEC and HCBECs were both able to induce memory (CD45RO ) but not naive (CD45RA ) CD4 T cells to secrete IFN- and IL-2. By CFSE dilution and FACS analysis, purified memory CD4 T cells, but not naive CD4 T cells, proliferated in cocultures with IFN- pretreated HUVECs and HCBECs (Fig. 5C). Similarly, only memory CD8 T cells were induced to secrete IFN- and IL-2 (Fig. 5, D and E) and to proliferate significantly (Fig. 5F) in response to allogeneic HUVECs or HCBECs, but in this case IFN- pretreatment was not required. The overall magnitude of T cell responses FIGURE 5. HUVECs and HCBECs activate CD4 and CD8 T cells memory but not naive T cells. A C, HUVECs and HCBECs were treated with IFN- as previously indicated and subsequently cocultured with CD4 T memory (CD45RO ) or naive (CD45RA ) T cells. A and B, Media were collected for ELISA from HUVECs and HCBECs cultured with either memory or naive CD4 T cells for 24 h. Memory CD4 T cells cultured with both HUVECs and HCBECs showed significant production of both IFN- (A) and IL-2 (B) when compared with naive CD4 T or when compared with resting ECs (not pretreated with IFN- ) or memory and naive CD4 T cells cultured in the absence of ECs. Data are mean SEM of quadruplicate determinations from one of two independent experiments with similar results. C, After 1 wk of cocultures, purified CFSE-labeled CD4 memory and naive T cells were stained with anti-cd4 mab and subjected to FACS analysis. Purified memory CD4 T cells showed significantly greater proliferation than their purified naive counterparts in HUVECs and HCBECs cocultures. Data are from one of two independent experiments with similar results. D F, HUVECs and HCBECs were cocultured with CD8 memory (CD45RO ) or naive (CD45RA ) T cells. D and E, Media were collected for ELISA from HUVECs and HCBECs cultured either memory or naive CD8 T cells for 48 h. Memory CD8 T cells cultured with both HUVECs and HCBECs showed significant production of both IFN- (A) and IL-2 (B) when compared with naive CD8 T cells or memory and naive CD8 T cells cultured in the absence of ECs. Data are mean SEM of quadruplicates determinations from one of two independent experiments with similar results. E, After 1 wk of cocultures, CFSE-labeled T cells were stained with anti-cd8 mab and subjected to FACS analysis. Purified memory CD8 T cells showed greater proliferation than their purified naive counterparts in HUVECs and HCBECs cocultures. Data are from one of two independent experiments with similar results.

7494 ALLOIMMUNITY TO ENDOTHELIUM FROM CORD BLOOD PROGENITORS to HCBECs were somewhat less than those elicited by allogeneic HUVECs, but differences did not reach statistical significance in multiple comparisons. To determine whether the variably lesser degree of stimulation induced by HCBECs from donor I correlated with the slightly higher level of expression of the IFN- -induced IDO (Fig. 2D), we either inhibited IDO activity with 1-methyl- (D)-tryptophan (200 M) or supplemented the cocultures with tryptophan (200 M). Neither of these strategies increased the ability of HCBECs (or HUVECs) to further stimulate T cell proliferation (data not shown). FIGURE 6. HUVECs and HCBECs are subjected to T cell-mediated rejection in vivo. A, H&E staining of collagen-fibronectin gels containing either HCBECs or HUVECs. Stable, perfused, microvessels were formed at 3 wk by either HCBECs or HUVECs in protein gels following s.c. implantation in immunodeficient mice (left panels) and were found to persist for at least 6 wk (middle panels). Three weeks after implantation, the remaining animals were reconstituted with PBMCs or given saline. Three weeks after reconstitution, vessel destruction was noted in gels containing either HCBECs or HUVECs in animals receiving PBMCs (right panels), but not saline injected controls (central panels). B, Quantification of human vessel density, calculated as Uea-1 positive staining per gel area in arbitrary units. For each animal, 2 4 randomly selected fields ( 20) from two sections were quantified; n 3., p 0.05 Functional analyses in vivo In a final series of experiments, we evaluated the interactions between allogeneic lymphocytes and human EC-lined microvessels in vivo using a previously described model. Specifically, human ECs were suspended in proteins gels formed from rat tail type I collagen and human plasma fibronectin (5, 36, 37) and then implanted into the abdominal wall of C.B-17/SCID-beige mouse. Consistent with previous reports, implanted HUVECs and HCBECs will form tubes that inosculate with the host vasculature and are perfused with mouse blood (Fig. 6A) (5, 14). Perfused tubes are established by 3 wk and persist for as long as 6 wk (Fig. 6A). To detect T cell-mediated alloreactivity in vivo, 3 wk after implantation of collagen-fibronectin gels containing HUVECs or HCBECs, some of the animals were inoculated i.p. with 3 10 8 PBMC allogeneic to the HUVECs or HCBECs lining the developed human microvessels. The remaining animals were injected with saline as control groups. All grafts were harvested 3 wk later. Grafts containing either HCBECs or HUVECs showed a plexus of simple EC-lined tubes containing mouse blood at 3 wk postimplantation and the appearance at 6 wk was largely unchanged in animals inoculated with saline (Fig. 6A, left and center panels). Inoculation with PBMCs led similar changes in grafts formed from HCBECs or HUVECs, namely loss of tubes, a sparse inflammatory infiltrate comprised of mononuclear cells, and focal areas of calcification (Fig. 6A, right panels). Immunostaining identified only rare human CD45 cells within the gel, although focal accumulations were noted at the edge of the gel in some specimens (data not shown). Staining of human ECs with Uea-1 lectin confirmed previous results with HUVECs that the perfused tubes within the gels of control animals were lined with human ECs and that human ECs disappeared in animals receiving allogeneic human PBMCs (5, 36, 37). Quantitation of Uea-1 staining revealed that the number of human EC-lined tubes increased between 3 and 6 wk in saline inoculated animals, but were decreased in animals receiving PBMCs (Fig. 6B). Significantly, no differences were observed in this in vivo model of human graft rejection between gels containing HCBECs or gels containing HUVECs. Discussion Previous reports have shown that human EPCs isolated from peripheral blood or umbilical cord blood can be differentiated to mature ECs and expanded in vitro, providing a robust source of ECs for cell transplantation and for promoting vascularization of tissue engineered organs and tissues (38 44). In theory, the expansion of vascular cells from blood progenitors could allow patients to serve as their own (autologous) stem cell donors, minimizing the potential for immunological rejection. However, most indications for the use of allogeneic sources of ECs will of necessity be derived from allogeneic stem cell sources, e.g., in the treatment of acute ischemic events or in the prefabrication of bioengineered tissues for clinical use. Moreover, cord blood EPCs appear more robust than EPCs from peripheral blood of adults and most adult recipients will not have access to autologous cord blood. Differentiation and expansion of ECs from allogeneic EPCs will give rise to alloantigenic differences between donor and recipient. Before this study, the significance of such differences for graft survival was unknown. It has been demonstrated that differentiated ECs isolated from mature blood vessels, such as HUVECs, are immunogenic,

The Journal of Immunology 7495 stimulating nonself (allogeneic) memory T cells to proliferate, differentiate, and produce effector cytokines (25, 35). Remarkably, there are no previous data comparing the capacities of circulating EPC-derived ECs with their counterparts isolated from mature blood vessels with regard to stimulating alloreactive T cells (45), the principal mediators of allograft rejection. The present study was conducted to address this question. We (10) and others (14, 16, 21) have demonstrated that late outgrowth EPCs from cord blood efficiently differentiate into ECs, and HCBECs appear to behave in a manner, except from their greater proliferative potential, that is indistinguishable from ECs isolated from a vessel wall such as HUVECs assessed by morphology, lineage markers, and tube formation capacity in vivo. In the present study, we show that HCBECs express essentially the same pattern of immunologically relevant molecules as HUVECs, including class I and class II MHC molecules, costimulators, adhesion molecules, cytokines, and chemokines under basal conditions and after cytokine stimulation. We also found that HCBECs are generally comparable to HUVECs in their ability to activate allogeneic memory CD4 and CD8 T cells, assessed by cytokine production and proliferation, although HCBECs may be slightly less efficient than HUVECs at eliciting these T cell responses in vitro. The IDO-mediated immunoinhibitory response was not responsible for this difference. We confirm that HCBECs possess a similar capacity as HUVECs to generate perfused human microvessel-like structures when incorporated in a collagen-fibronectin gel and implanted in the s.c. position on the abdominal wall of immunocompromised mice. We did confirm that when mice that had received microvascular grafts prepared from HUVECs were reconstituted with human PBMCs allogeneic to the cells used to generate the human microvascular tissue, the vessels were subjected to PBMC-mediated vessel destruction (37). Unpublished studies (J. S. Pober and A. L. M. Bothwell) using T cells primed by coculture with allogeneic ECs suggested that vessel destruction in this model is likely the result of an allogeneic response. In the present study, we found that vessel destruction was indistinguishable when the grafts were constructed with HCBECs or with HUVECs. Cumulatively, these findings establish that ECs differentiated from EPCs are immunogenic and indicate that they will likely be rejected by the immune system if used for cell therapy and/or tissue engineering. Our findings suggest that EPCs will need to be obtained from MHC-matched donors to reduce alloantigenicity. Alternatively, it may be possible to genetically alter such cells to reduce their capacity to activate T cells. Acknowledgments We thank Louise Benson, Gwen Davis, and Lisa Gras for assistance with cell culture and animal care, and Dr. Edward Snyder and the Yale Blood Bank for assistance with leukaphereses. Disclosures The authors have no financial conflict of interest. References 1. Carmeliet, P. 2005. Angiogenesis in life, disease and medicine. Nature 438: 932 936. 2. Jain, R. K. 2003. Molecular regulation of vessel maturation. Nat. Med. 9: 685 693. 3. Folkman, J. 1995. Angiogenesis in cancer, vascular, rheumatoid and other disease. Nat. Med. 1: 27 31. 4. Conway, E. M., D. Collen, and P. Carmeliet. 2001. Molecular mechanisms of blood vessel growth. Cardiovasc. Res. 49: 507 521. 5. Schechner, J. S., A. K. Nath, L. Zheng, M. S. Kluger, C. C. Hughes, M. R. Sierra-Honigmann, M. I. Lorber, G. Tellides, M. Kashgarian, A. L. Bothwell, and J. S. Pober. 2000. In vivo formation of complex microvessels lined by human endothelial cells in an immunodeficient mouse. Proc. Natl. Acad. Sci. USA 97: 9191 9196. 6. Enis, D. R., B. R. Shepherd, Y. Wang, A. Qasim, C. M. Shanahan, P. L. Weissberg, M. Kashgarian, J. S. Pober, and J. S. Schechner. 2005. Induction, differentiation, and remodeling of blood vessels after transplantation of Bcl- 2-transduced endothelial cells. Proc. Natl. Acad. Sci. USA 102: 425 430. 7. Garfinkel, S., S. Brown, J. H. Wessendorf, and T. Maciag. 1994. Post-transcriptional regulation of interleukin 1 in various strains of young and senescent human umbilical vein endothelial cells. Proc. Natl. Acad. Sci. USA 91: 1559 1563. 8. Asahara, T., T. Murohara, A. Sullivan, M. Silver, R. van der Zee, T. Li, B. Witzenbichler, G. Schatteman, and J. M. Isner. 1997. Isolation of putative progenitor endothelial cells for angiogenesis. Science 275: 964 967. 9. Kaushal, S., G. E. Amiel, K. J. Guleserian, O. M. Shapira, T. Perry, F. W. Sutherland, E. Rabkin, A. M. Moran, F. J. Schoen, A. Atala, et al. 2001. Functional small-diameter neovessels created using endothelial progenitor cells expanded ex vivo. Nat. Med. 7: 1035 1040. 10. Shepherd, B. R., D. R. Enis, F. Wang, Y. Suarez, J. S. Pober, and J. S. Schechner. 2006. Vascularization and engraftment of a human skin substitute using circulating progenitor cell-derived endothelial cells. FASEB J. 20: 1739 1741. 11. Zhao, Y., D. Glesne, and E. Huberman. 2003. A human peripheral blood monocyte-derived subset acts as pluripotent stem cells. Proc. Natl. Acad. Sci. USA 100: 2426 2431. 12. Schmeisser, A., C. D. Garlichs, H. Zhang, S. Eskafi, C. Graffy, J. Ludwig, R. H. Strasser, and W. G. Daniel. 2001. Monocytes coexpress endothelial and macrophagocytic lineage markers and form cord-like structures in Matrigel under angiogenic conditions. Cardiovasc. Res. 49: 671 680. 13. Fernandez Pujol, B., F. C. Lucibello, U. M. Gehling, K. Lindemann, N. Weidner, M. L. Zuzarte, J. Adamkiewicz, H. P. Elsasser, R. Muller, and K. Havemann. 2000. Endothelial-like cells derived from human CD14 positive monocytes. Differentiation 65: 287 300. 14. Yoder, M. C., L. E. Mead, D. Prater, T. R. Krier, K. N. Mroueh, F. Li, R. Krasich, C. J. Temm, J. T. Prchal, and D. A. Ingram. 2007. Redefining endothelial progenitor cells via clonal analysis and hematopoietic stem/progenitor cell principals. Blood 109: 1801 1809. 15. Rehman, J., J. Li, C. M. Orschell, and K. L. March. 2003. Peripheral blood endothelial progenitor cells are derived from monocyte/macrophages and secrete angiogenic growth factors. Circulation 107: 1164 1169. 16. Ingram, D. A., N. M. Caplice, and M. C. Yoder. 2005. Unresolved questions, changing definitions, and novel paradigms for defining endothelial progenitor cells. Blood 106: 1525 1531. 17. Ingram, D. A., L. E. Mead, D. B. Moore, W. Woodard, A. Fenoglio, and M. C. Yoder. 2005. Vessel wall-derived endothelial cells rapidly proliferate because they contain a complete hierarchy of endothelial progenitor cells. Blood 105: 2783 2786. 18. Ingram, D. A., L. E. Mead, H. Tanaka, V. Meade, A. Fenoglio, K. Mortell, K. Pollok, M. J. Ferkowicz, D. Gilley, and M. C. Yoder. 2004. Identification of a novel hierarchy of endothelial progenitor cells using human peripheral and umbilical cord blood. Blood 104: 2752 2760. 19. Hur, J., C. H. Yoon, H. S. Kim, J. H. Choi, H. J. Kang, K. K. Hwang, B. H. Oh, M. M. Lee, and Y. B. Park. 2004. Characterization of two types of endothelial progenitor cells and their different contributions to neovasculogenesis. Arterioscler. Thromb. Vasc. Biol. 24: 288 293. 20. Gulati, R., D. Jevremovic, T. E. Peterson, S. Chatterjee, V. Shah, R. G. Vile, and R. D. Simari. 2003. Diverse origin and function of cells with endothelial phenotype obtained from adult human blood. Circ. Res. 93: 1023 1025. 21. Bompais, H., J. Chagraoui, X. Canron, M. Crisan, X. H. Liu, A. Anjo, C. Tolla-Le Port, M. Leboeuf, P. Charbord, A. Bikfalvi, and G. Uzan. 2004. Human endothelial cells derived from circulating progenitors display specific functional properties compared with mature vessel wall endothelial cells. Blood 103: 2577 2584. 22. Epperson, D. E., and J. S. Pober. 1994. Antigen-presenting function of human endothelial cells: direct activation of resting CD8 T cells. J. Immunol. 153: 5402 5412. 23. Marelli-Berg, F. M., L. Frasca, L. Weng, G. Lombardi, and R. I. Lechler. 1999. Antigen recognition influences transendothelial migration of CD4 T cells. J. Immunol. 162: 696 703. 24. Shiao, S. L., J. M. McNiff, and J. S. Pober. 2005. Memory T cells and their costimulators in human allograft injury. J. Immunol. 175: 4886 4896. 25. Ma, W., and J. S. Pober. 1998. Human endothelial cells effectively costimulate cytokine production by, but not differentiation of, naive CD4 T cells. J. Immunol. 161: 2158 2167. 26. Beutelspacher, S. C., P. H. Tan, M. O. McClure, D. F. Larkin, R. I. Lechler, and A. J. George. 2006. Expression of 2,3-dioxygenase (IDO) by endothelial cells: implications for the control of alloresponses. Am. J. Transplant. 6: 1320 1330. 27. Gimbrone, M. A., Jr., R. S. Cotran, and J. Folkman. 1974. Human vascular endothelial cells in culture: growth and DNA synthesis. J. Cell Biol. 60: 673 684. 28. Kirkiles-Smith, N. C., K. Mahboubi, J. Plescia, J. M. McNiff, J. Karras, J. S. Schechner, D. C. Altieri, and J. S. Pober. 2004. IL-11 protects human microvascular endothelium from alloinjury in vivo by induction of survivin expression. J. Immunol. 172: 1391 1396. 29. Bender, J. R., M. M. Sadeghi, C. Watson, S. Pfau, and R. Pardi. 1994. Heterogeneous activation thresholds to cytokines in genetically distinct endothelial cells: evidence for diverse transcriptional responses. Proc. Natl. Acad. Sci. USA 91: 3994 3998. 30. Pober, J. S., T. Collins, M. A. Gimbrone, Jr., R. S. Cotran, J. D. Gitlin, W. Fiers, C. Clayberger, A. M. Krensky, S. J. Burakoff, and C. S. Reiss. 1983. Lymphocytes recognize human vascular endothelial and dermal fibroblast Ia antigens induced by recombinant immune interferon. Nature 305: 726 729.

7496 ALLOIMMUNITY TO ENDOTHELIUM FROM CORD BLOOD PROGENITORS 31. Collins, T., L. A. Lapierre, W. Fiers, J. L. Strominger, and J. S. Pober. 1986. Recombinant human tumor necrosis factor increases mrna levels and surface expression of HLA-A,B antigens in vascular endothelial cells and dermal fibroblasts in vitro. Proc. Natl. Acad. Sci. USA 83: 446 450. 32. Karmann, K., C. C. Hughes, J. Schechner, W. C. Fanslow, and J. S. Pober. 1995. CD40 on human endothelial cells: inducibility by cytokines and functional regulation of adhesion molecule expression. Proc. Natl. Acad. Sci. USA 92: 4342 4346. 33. Kurt-Jones, E. A., W. Fiers, and J. S. Pober. 1987. Membrane interleukin 1 induction on human endothelial cells and dermal fibroblasts. J. Immunol. 139: 2317 2324. 34. Mellor, A. L., and D. H. Munn. 2004. IDO expression by dendritic cells: tolerance and tryptophan catabolism. Nat. Rev. Immunol. 4: 762 764. 35. Dengler, T. J., D. R. Johnson, and J.S. Pober. 2001. Human vascular endothelial cells stimulate a lower frequency of alloreactive CD8 pre-ctl and induce less clonal expansion than matching B lymphoblastoid cells: development of a novel limiting dilution analysis method based on CFSE labeling of lymphocytes. J. Immunol. 166: 3846 3854. 36. Zheng, L., T. J. Dengler, M. S. Kluger, L. A. Madge, J. S. Schechner, S. E. Maher, J. S. Pober, and A. L. Bothwell. 2000. Cytoprotection of human umbilical vein endothelial cells against apoptosis and CTL-mediated lysis provided by caspase-resistant Bcl-2 without alterations in growth or activation responses. J. Immunol. 164: 4665 4671. 37. Zheng, L., T. F. Gibson, J. S. Schechner, J. S. Pober, and A. L. Bothwell. 2004. Bcl-2 transduction protects human endothelial cell synthetic microvessel grafts from allogeneic T cells in vivo. J. Immunol. 173: 3020 3026. 38. Griese, D. P., A. Ehsan, L. G. Melo, D. Kong, L. Zhang, M. J. Mann, R. E. Pratt, R. C. Mulligan, and V. J. Dzau. 2003. Isolation and transplantation of autologous circulating endothelial cells into denuded vessels and prosthetic grafts: implications for cell-based vascular therapy. Circulation 108: 2710 2715. 39. Assmus, B., V. Schachinger, C. Teupe, M. Britten, R. Lehmann, N. Dobert, F. Grunwald, A. Aicher, C. Urbich, H. Martin, et al. 2002. Transplantation of progenitor cells and regeneration enhancement in acute myocardial infarction (TOPCARE-AMI). Circulation 106: 3009 3017. 40. Kalka, C., H. Masuda, T. Takahashi, W. M. Kalka-Moll, M. Silver, M. Kearney, T. Li, J. M. Isner, and T. Asahara. 2000. Transplantation of ex vivo expanded endothelial progenitor cells for therapeutic neovascularization. Proc. Natl. Acad. Sci. USA 97: 3422 3427. 41. Kawamoto, A., H. C. Gwon, H. Iwaguro, J. I. Yamaguchi, S. Uchida, H. Masuda, M. Silver, H. Ma, M. Kearney, J. M. Isner, and T. Asahara. 2001. Therapeutic potential of ex vivo expanded endothelial progenitor cells for myocardial ischemia. Circulation 103: 634 637. 42. Kocher, A. A., M. D. Schuster, M. J. Szabolcs, S. Takuma, D. Burkhoff, J. Wang, S. Homma, N. M. Edwards, and S. Itescu. 2001. Neovascularization of ischemic myocardium by human bone-marrow-derived angioblasts prevents cardiomyocyte apoptosis, reduces remodeling and improves cardiac function. Nat. Med. 7: 430 436. 43. Pesce, M., A. Orlandi, M. G. Iachininoto, S. Straino, A. R. Torella, V. Rizzuti, G. Pompilio, G. Bonanno, G. Scambia, and M. C. Capogrossi. 2003. Myoendothelial differentiation of human umbilical cord blood-derived stem cells in ischemic limb tissues. Circ. Res. 93: e51 e62 44. Shirota, T., H. Yasui, H. Shimokawa, and T. Matsuda. 2003. Fabrication of endothelial progenitor cell (EPC)-seeded intravascular stent devices and in vitro endothelialization on hybrid vascular tissue. Biomaterials 24: 2295 2302. 45. Fang, T. C., M. R. Alison, N. A. Wright, and R. Poulsom. 2004. Adult stem cell plasticity: will engineered tissues be rejected? Int. J. Exp. Pathol. 85: 115 124.