- Email: [email protected]
Vascular endothelial differentiation in sponge matrix allografts
Vascular Endothelial Differentiation in Sponge Matrix Allografts D. Keith Bishop, Daniel D. Sedmak, Douglas M. Leppink, and Charles G. Orosz
ABSTRACT: These studies test the hypothesis that vascular endothelia in sponge allografts may develop a function and phenotype resembling the high endothelial venules (HEV) in lymph nodes, thus facilitating the lymphocytic infiltration that is characteristic of allograft rejection, Using limiting dilution analysis to quantitate helper-T-cell accumulation at graft sites, immunohistologic analysis of graft reactivity with the HEV-specific monoclonal antibody MECA 325, and ex vivo lymphocyteendothelial adhesion assays with graft tissues, we obtained evidence to suggest that HEV-like endothelia may develop at a graft site but that the process whereby lymphocytes accumulate at a graft site is more complex than was initially expected. ABBREVIATIONS HEV high endothelial venules HTL helper T cells LDA limitingdilution analysis
LN mAb
lymph nodes monoclonalantibody
INTRODUCTION Acute allograft rejection is characterized by lymphocytic infiltration of the graft site [1-3], but the mechanisms which govern this phenomenon are not well understood. As part of the process of homeostatic lymphocyte recirculation, blood-borne iymphocytes bind to and traverse the specialized high endothelial venules (HEV) that are constitutively present in peripheral lymphoid tissues, thereby entering the lymphatics [4-7]. The studies outlined in this report were designed to test the hypothesis that specialized endothelia similar to lymph node HEV develop at a neovascularized graft site and mediate the lymphocytic infiltration associated with allograft rejection. MATERIALS A N D METHODS Sponge allografts. Female C57B1/6 mice were implanted subcutaneously with
paired polyurethane sponge cylinders (8 x 15 mm) and injected with allogeneic (DBA/2) or syngeneic splenocytes (3.5 million cells per sponge) as previously From the University of Utah Schoolof Medicine, Division of Hematology/Oncology, Salt Lake City, Utah (D.K.B.); the Ohio State University Collegeof Medicine, Department~f Pathology, Columbus, Ohio (D.D.S.); the Ohio State University College of Medicine, Department of Surgery, Columbus, Ohio (D.M.L.); and the Ohio State University Comprehensive Cancer Center, Columbus, Ohio (C.G.D.). Address reprint requests to D. Keith Bishop, Ph.D., University of Utah School of Medicine, Division of Hematology/Oncology, Salt Lake City, UT 84132. Received December I, 1989: acceptedJanuary 3, 1990.
128 0198-8859/90/$3.50
Human Immunology 28, 128-133 (1990) © American Society for Histocompatibility and Immunogenetics, 1990
HEY-like Endothelia in Sponge Grafts
129
TABLE 1 Distribution of donor alloantigen-reactive HTL in sponge allograft vs. sponge isograft recipients
Allograft Isograft
Site
# H T L / 1 0 6 cells~
C.I.
Graft Peripheral blood Graft P e r i p h e r a l blood
260 163 N.D. b 124
181-339 114-213 -87-163
Minimal frequency estimate obtained by chi square minimization; C.I. = 95 % confidence interval for analysis of response in limiting dilution microcultures. h N.D. = not detectable
described [8]. Relevant tissues were harvested 10 days following sponge implantation for further study.
Limiting dilution analysis (LDA)." The number of DBA/2-reactive helper T cells (HTL) present in various tissues of sponge allograft-and sponge isograft-bearing mice was quantitated using LDA techniques described in detail elsewhere [9].
Immunohistologic analysis. The monoclonal antibody (mAb) MECA 325 (rat IgG2a) specifically reacts with murine HEV [10] and was kindly provided by Dr. Eugene Butcher at Stanford University. Frozen 10-/.Lmsections of sponge allograft and isograft tissues were reacted with mAb MECA 325 at predetermined optimal concentrations. Antibody reactivity was visualized by standard immunoperoxidase techniques using horseradish peroxidase-conjugated rabbit anti-rat IgG and subsequent development with 3-amino-9-ethylcarbazole and hydrogen peroxide.
Lymphocyte-endothelialadhesion assay. Frozen 12-/zm sections of sponge allograftor isograft-associated tissues were overlayed with 5 × 105 mesenteric lymph node cells for 20 min, fixed overnight in 1.5 % glutaraldehyde, and counterstained with 0.1% thionin [11,12]. RESULTS As shown in Table 1, sponge allografts, but not sponge isografts, acquire LDAdetectable alloreactive HTL, despite the fact that these HTL are always present in the peripheral blood of both allograft and isograft recipients. To explain these observations, we hypothesized that the vascular endothelia of sponge isografrs maintain a phenotype characteristic of noninflamed tissues and remain a barrier to LDA-detectable T cells, while the vascular endothelia in sponge allografts develop a phenotype similar to lymph node HEV and allow the extravasation of LDA-detectable T cells. To test this, we performed immunohistologic analyses of graft tissues with a mAb called MECA 325, which reacts selectively with murine lymph node HEV [10]. We observed that sponge allografts contain numerous vessels reactive with mAb MECA 325 [Figure 1, (A)], many of which were associated with perivascular cuffs of lymphocytes. However, sponge isografts also contained MECA 325-reactive vessels [Figure 1, (B)]. Although the numbers of MECA 325-positive vessels present in sponge isografts or allografts were similar (data not shown), computer-assisted morphometric analysis of 300 MECA 325-positive vessels present in sponge allografts and in sponge isografts revealed a highly significant
130
A
F I G U R E 1 Reactivity of vascular structures in sponge allografts (A) and sponge isografts (B) with mAb MECA 325. Arrows indicate reactivity as revealed by indirect immunoperoxidase techniques on frozen tissue sections ( x 63).
HEV-like Endothelia in Sponge Grafts TABLE 2
131
Morphometric analysis of MECA 325-positive vessels in sponge allografts and isografts Sponge allograft
Mean vessel perimeter,/.Lm (S.D.) Mean luminal area, ~m 2 (S.D.)
Sponge isograft
p
373
176
~0.0001
(242) 6810
(91) 1784
~0,0001
(8069)
(216)
difference between both the mean perimeter and luminal areas of MECA 325-positive vessels (Table 2). As a further step in this analysis, we employed ex vivo lymphocyte-endothelia adhesion assays to determine whether vessels in sponge implants were capable of interacting with lymphocytes. We observed that certain of the blood vessels in sponge allografts readily bound mesenteric LN cells in this assay [Figure 2 (A)] while blood vessels in sponge isografts did not [Figure 2 (B)].
DISCUSSION We have tested the hypothesis that vascular endothelia in allografts differentiate to resemble LN HEV, thus allowing the lymphocyte infiltration characteristic of allograft rejection. We observed experimentally that (1) sponge ailografts, but not sponge isografts, acquire LDA-detectable, alloreactive HTL, (2) both allografts and isografts develop HEV-like, MECA 325-reactive vascular endothelia, although there is a morphologic difference between the MECA 325-reactive vessels in these two tissues, and (3) sponge allograft, but not sponge isografts, contain vascular endothelia with a detectable affinity for lymphocytes. Hence, HEV-like MECA 325-reactive vessels develop in sponge allografts regardless of the presence or absence of alloantigens. Further, the presence of MECA 325-reactive vessels does not directly correlate with the accumulation of alloreactive HTL nor with the functional capacity of vascular structures to bind lymphocytes. Thus, LN-like lymphocyte-endothelial reactions may develop at a graft site, but this process is more complex than was initially anticipated. The fact that the MECA 325-reactive vessels in isografts apparently do not function for lymphocyte adhesion suggests that the vascular endothelia in rejecting aliografts may undergo staged differentiation: (1) development of vessels reactive with HEV-specific mAbs and (2) the acquisition of as yet undefined properties which allow the vessels to mediate lymphocyte adhesion and extravasation. This final stage of endothelial differentiation would likely be regulated by the alloantigendriven cytokine production [13-15] present in allografts but not isografts. In general, our findings illustrate the utility of sponge implants as an in vivo model for evaluating the mechanisms involved in lymphocytic infiltration into rejecting allografts. I t will be of interest to determine how pharmacologic intervention influences endothelial behavior in this experimental system. The concept of interfering with lymphocyte-endothelial interactions in vivo provides a new strategy for immunosuppressive therapies aimed at prolongation of allograft survival.
132
A
B
F I G U R E 2 Ability of vascular structures in sponge allografts (A) but not sponge isografts (B) to bind mesenteric LN lymphocytes in ex vivo adhesion assays. Arrows denote vascular structures in sponge isografts (x 63).
HEV-like Endothelia in Sponge Grafts
133
ACKNOWLEDGMENTS We wish to thank Ms. Cindy Narcross for her secretarial assistance and Ms. Teri Bailey for her administrative assistance during these studies. This is paper 48 from the Therapeutic Immunology Laboratories at Ohio State University. This publication was supported by N I H grants R01-A124676 and R01-DK34774 and PHS grant P 30 CA1605814 awarded by the National Cancer Institute, Department of Health and Human Services.
REFERENCES 1. Strom TB, Tilney N, Paradysz JM, Bancewicz J, Carpenter CB: J Immunol 118:2020, 1977. 2. von Willebrand E, Soots A, Hayry P: Cell Immunol 46:309, 1979. 3. Renkonen R, Soots A, von Willebrand E, Hayry P: Cell Immunol 77:187, 1983. 4. Gowans JL, Knight EJ: Proc R Soc Lond Set B 159:257, 1964. 5. Jalkanen S, Reichert RA, Gallatin WM, Bargatze RF, Weissman IL, Butcher EC: Immunol Rev 91:39, 1986. 6. WoodruffJJ, Clarke LM, Chin YH: Annu Rev Immunol 5:201, 1987. 7. Berg EL, Goldstein LA, Jutila MA, Nakache M, Picker LJ, Streeter PR, Wu NW, Zhou D, Butcher EC: Immunol Rev 108:5, 1989. 8. Orosz CG, Zinn NE, Sirinek L, Ferguson RM: Transplantation 41:75, 1986. 9. Bishop DK, Orosz CG: Transplantation 47:67 I, 1989. 10. Duijvestijin AM, Kerkhove M, Bargatze RF, Butcher EC: J Immunol 138:713, 1987. 11. Stamper HB Jr, WoodruffJJ: J Exp Med 144:828, 1976. 12. Jalkanen ST, Butcher EC: Blood 66:577, 1985. 13. Dumonde DC, Pulley MS, Paradinas FJ, Southcott BS, O'Connel D, Robinson MRG, den Hollander F, Schuurs AH: J Pathol 138:289, 1982. 14. Polverimi PJ, Cotran RS, Sholley MM: J Immunol 118:529, 1977. 15. Hanto DW, Harry JT, Hoffman RA, Simmons RL: Transplantation 42:621, 1986.
ABSTRACT: These studies test the hypothesis that vascular endothelia in sponge allografts may develop a function and phenotype resembling the high endothelial venules (HEV) in lymph nodes, thus facilitating the lymphocytic infiltration that is characteristic of allograft rejection, Using limiting dilution analysis to quantitate helper-T-cell accumulation at graft sites, immunohistologic analysis of graft reactivity with the HEV-specific monoclonal antibody MECA 325, and ex vivo lymphocyteendothelial adhesion assays with graft tissues, we obtained evidence to suggest that HEV-like endothelia may develop at a graft site but that the process whereby lymphocytes accumulate at a graft site is more complex than was initially expected. ABBREVIATIONS HEV high endothelial venules HTL helper T cells LDA limitingdilution analysis
LN mAb
lymph nodes monoclonalantibody
INTRODUCTION Acute allograft rejection is characterized by lymphocytic infiltration of the graft site [1-3], but the mechanisms which govern this phenomenon are not well understood. As part of the process of homeostatic lymphocyte recirculation, blood-borne iymphocytes bind to and traverse the specialized high endothelial venules (HEV) that are constitutively present in peripheral lymphoid tissues, thereby entering the lymphatics [4-7]. The studies outlined in this report were designed to test the hypothesis that specialized endothelia similar to lymph node HEV develop at a neovascularized graft site and mediate the lymphocytic infiltration associated with allograft rejection. MATERIALS A N D METHODS Sponge allografts. Female C57B1/6 mice were implanted subcutaneously with
paired polyurethane sponge cylinders (8 x 15 mm) and injected with allogeneic (DBA/2) or syngeneic splenocytes (3.5 million cells per sponge) as previously From the University of Utah Schoolof Medicine, Division of Hematology/Oncology, Salt Lake City, Utah (D.K.B.); the Ohio State University Collegeof Medicine, Department~f Pathology, Columbus, Ohio (D.D.S.); the Ohio State University College of Medicine, Department of Surgery, Columbus, Ohio (D.M.L.); and the Ohio State University Comprehensive Cancer Center, Columbus, Ohio (C.G.D.). Address reprint requests to D. Keith Bishop, Ph.D., University of Utah School of Medicine, Division of Hematology/Oncology, Salt Lake City, UT 84132. Received December I, 1989: acceptedJanuary 3, 1990.
128 0198-8859/90/$3.50
Human Immunology 28, 128-133 (1990) © American Society for Histocompatibility and Immunogenetics, 1990
HEY-like Endothelia in Sponge Grafts
129
TABLE 1 Distribution of donor alloantigen-reactive HTL in sponge allograft vs. sponge isograft recipients
Allograft Isograft
Site
# H T L / 1 0 6 cells~
C.I.
Graft Peripheral blood Graft P e r i p h e r a l blood
260 163 N.D. b 124
181-339 114-213 -87-163
Minimal frequency estimate obtained by chi square minimization; C.I. = 95 % confidence interval for analysis of response in limiting dilution microcultures. h N.D. = not detectable
described [8]. Relevant tissues were harvested 10 days following sponge implantation for further study.
Limiting dilution analysis (LDA)." The number of DBA/2-reactive helper T cells (HTL) present in various tissues of sponge allograft-and sponge isograft-bearing mice was quantitated using LDA techniques described in detail elsewhere [9].
Immunohistologic analysis. The monoclonal antibody (mAb) MECA 325 (rat IgG2a) specifically reacts with murine HEV [10] and was kindly provided by Dr. Eugene Butcher at Stanford University. Frozen 10-/.Lmsections of sponge allograft and isograft tissues were reacted with mAb MECA 325 at predetermined optimal concentrations. Antibody reactivity was visualized by standard immunoperoxidase techniques using horseradish peroxidase-conjugated rabbit anti-rat IgG and subsequent development with 3-amino-9-ethylcarbazole and hydrogen peroxide.
Lymphocyte-endothelialadhesion assay. Frozen 12-/zm sections of sponge allograftor isograft-associated tissues were overlayed with 5 × 105 mesenteric lymph node cells for 20 min, fixed overnight in 1.5 % glutaraldehyde, and counterstained with 0.1% thionin [11,12]. RESULTS As shown in Table 1, sponge allografts, but not sponge isografts, acquire LDAdetectable alloreactive HTL, despite the fact that these HTL are always present in the peripheral blood of both allograft and isograft recipients. To explain these observations, we hypothesized that the vascular endothelia of sponge isografrs maintain a phenotype characteristic of noninflamed tissues and remain a barrier to LDA-detectable T cells, while the vascular endothelia in sponge allografts develop a phenotype similar to lymph node HEV and allow the extravasation of LDA-detectable T cells. To test this, we performed immunohistologic analyses of graft tissues with a mAb called MECA 325, which reacts selectively with murine lymph node HEV [10]. We observed that sponge allografts contain numerous vessels reactive with mAb MECA 325 [Figure 1, (A)], many of which were associated with perivascular cuffs of lymphocytes. However, sponge isografts also contained MECA 325-reactive vessels [Figure 1, (B)]. Although the numbers of MECA 325-positive vessels present in sponge isografts or allografts were similar (data not shown), computer-assisted morphometric analysis of 300 MECA 325-positive vessels present in sponge allografts and in sponge isografts revealed a highly significant
130
A
F I G U R E 1 Reactivity of vascular structures in sponge allografts (A) and sponge isografts (B) with mAb MECA 325. Arrows indicate reactivity as revealed by indirect immunoperoxidase techniques on frozen tissue sections ( x 63).
HEV-like Endothelia in Sponge Grafts TABLE 2
131
Morphometric analysis of MECA 325-positive vessels in sponge allografts and isografts Sponge allograft
Mean vessel perimeter,/.Lm (S.D.) Mean luminal area, ~m 2 (S.D.)
Sponge isograft
p
373
176
~0.0001
(242) 6810
(91) 1784
~0,0001
(8069)
(216)
difference between both the mean perimeter and luminal areas of MECA 325-positive vessels (Table 2). As a further step in this analysis, we employed ex vivo lymphocyte-endothelia adhesion assays to determine whether vessels in sponge implants were capable of interacting with lymphocytes. We observed that certain of the blood vessels in sponge allografts readily bound mesenteric LN cells in this assay [Figure 2 (A)] while blood vessels in sponge isografts did not [Figure 2 (B)].
DISCUSSION We have tested the hypothesis that vascular endothelia in allografts differentiate to resemble LN HEV, thus allowing the lymphocyte infiltration characteristic of allograft rejection. We observed experimentally that (1) sponge ailografts, but not sponge isografts, acquire LDA-detectable, alloreactive HTL, (2) both allografts and isografts develop HEV-like, MECA 325-reactive vascular endothelia, although there is a morphologic difference between the MECA 325-reactive vessels in these two tissues, and (3) sponge allograft, but not sponge isografts, contain vascular endothelia with a detectable affinity for lymphocytes. Hence, HEV-like MECA 325-reactive vessels develop in sponge allografts regardless of the presence or absence of alloantigens. Further, the presence of MECA 325-reactive vessels does not directly correlate with the accumulation of alloreactive HTL nor with the functional capacity of vascular structures to bind lymphocytes. Thus, LN-like lymphocyte-endothelial reactions may develop at a graft site, but this process is more complex than was initially anticipated. The fact that the MECA 325-reactive vessels in isografts apparently do not function for lymphocyte adhesion suggests that the vascular endothelia in rejecting aliografts may undergo staged differentiation: (1) development of vessels reactive with HEV-specific mAbs and (2) the acquisition of as yet undefined properties which allow the vessels to mediate lymphocyte adhesion and extravasation. This final stage of endothelial differentiation would likely be regulated by the alloantigendriven cytokine production [13-15] present in allografts but not isografts. In general, our findings illustrate the utility of sponge implants as an in vivo model for evaluating the mechanisms involved in lymphocytic infiltration into rejecting allografts. I t will be of interest to determine how pharmacologic intervention influences endothelial behavior in this experimental system. The concept of interfering with lymphocyte-endothelial interactions in vivo provides a new strategy for immunosuppressive therapies aimed at prolongation of allograft survival.
132
A
B
F I G U R E 2 Ability of vascular structures in sponge allografts (A) but not sponge isografts (B) to bind mesenteric LN lymphocytes in ex vivo adhesion assays. Arrows denote vascular structures in sponge isografts (x 63).
HEV-like Endothelia in Sponge Grafts
133
ACKNOWLEDGMENTS We wish to thank Ms. Cindy Narcross for her secretarial assistance and Ms. Teri Bailey for her administrative assistance during these studies. This is paper 48 from the Therapeutic Immunology Laboratories at Ohio State University. This publication was supported by N I H grants R01-A124676 and R01-DK34774 and PHS grant P 30 CA1605814 awarded by the National Cancer Institute, Department of Health and Human Services.
REFERENCES 1. Strom TB, Tilney N, Paradysz JM, Bancewicz J, Carpenter CB: J Immunol 118:2020, 1977. 2. von Willebrand E, Soots A, Hayry P: Cell Immunol 46:309, 1979. 3. Renkonen R, Soots A, von Willebrand E, Hayry P: Cell Immunol 77:187, 1983. 4. Gowans JL, Knight EJ: Proc R Soc Lond Set B 159:257, 1964. 5. Jalkanen S, Reichert RA, Gallatin WM, Bargatze RF, Weissman IL, Butcher EC: Immunol Rev 91:39, 1986. 6. WoodruffJJ, Clarke LM, Chin YH: Annu Rev Immunol 5:201, 1987. 7. Berg EL, Goldstein LA, Jutila MA, Nakache M, Picker LJ, Streeter PR, Wu NW, Zhou D, Butcher EC: Immunol Rev 108:5, 1989. 8. Orosz CG, Zinn NE, Sirinek L, Ferguson RM: Transplantation 41:75, 1986. 9. Bishop DK, Orosz CG: Transplantation 47:67 I, 1989. 10. Duijvestijin AM, Kerkhove M, Bargatze RF, Butcher EC: J Immunol 138:713, 1987. 11. Stamper HB Jr, WoodruffJJ: J Exp Med 144:828, 1976. 12. Jalkanen ST, Butcher EC: Blood 66:577, 1985. 13. Dumonde DC, Pulley MS, Paradinas FJ, Southcott BS, O'Connel D, Robinson MRG, den Hollander F, Schuurs AH: J Pathol 138:289, 1982. 14. Polverimi PJ, Cotran RS, Sholley MM: J Immunol 118:529, 1977. 15. Hanto DW, Harry JT, Hoffman RA, Simmons RL: Transplantation 42:621, 1986.