Volume 13 Number 5 May 1991
Specialcommunication
level of a single focal adhesion, groups of adhesion plaques, a single cell, or groups of cells. It should be possible to manipulate the composition of the extracellular matrix to quantitate the effects on cell adhesion. It should also be possible to use competitive binding studies of adhesion proteins, as well as selective blocking of domains of adhesion proteins by use of specific antibodies, to obtain insights concerning the molecular mechanisms of cell adhesion. Focal adhesions in endothelial cells may represent a complex of proteins capable of signaling the effects of external forces such as shear stress to the interior of the cell. It would be of interest to investigate whether such structures and their associated proteins play a regulatory role in the responses of endothelial cells to external physical forces. Such experiments are critical to an understanding of the adhesion of endothelial cells on vascular grafts where cells must resist detachment by flow and where the presence of an endothelial lining is likely to assist graft patency and survival. Peteev F. Davies, PhD Pritekw School of Medicine The University of Chicgo Chicago, Ill.
REFERENCES 1. Holtz J, Forstermann U, Pohl V, Giesler M, Bassenge EJ. Flow-dependent, endothelium-mediated dilation of epicardial coronary arteries in conscious dogs: effects of cycle-oxygenase inhibition. J Cardiovasc Pharmacol 1984;6:1161-9. 2. Langille BL, O’Donnell F. Reductions in arterial diameter produced by chronic decreases in blood flow are endothelialdependent. Science 1986;231:405-8. 3. Davies PF. Endothelial cells, hemodynamic forces, and the localization of atherosclerosis. In: Ryan US., ed. Endothelial Cells, Vol. II. Boca Baton, Florida: CRC Press, 1988: 123-39. 4. Olesen SP, Clapham DE, Davies PF. Hemodynamic shear stressactivates a K + current in vascular endothelial cells. Nature 1988;331:168-170. 5. Frangos JA, Eskin SE, McIntyre LV, Ives CL Flow effects on prostacyclin production by cultured human endothelial cells. Science 1985;227:1477-9. Diamond SL, McIntyre LV, Share&in JB, Dieffenbach C, Scott KF, Eskin SG. J Cell Physiol (In press). Davies PF. How do vascular endothelial cells respond to flow? News in Physiol Sci 1989;4:22-6. Dull RO, Davies PF. Hemodynamic shear stress modulates signal-response coupling in vascular endothelial cells (In press). Paddock SW. Tandem scanning reflected-light microscopy of cell-substratum adhesions and stress fibres in Swiss 3T3 cells. J Cell Sci 1989;93: 142-6.
ENDOTHELIAL
CELL SEEDING
Endothelial cell seeding is the transplantation of vascular endothelial cells to denuded vascular surfaces. Endothelium is a crucial cell in the modulation of clotting and inflammatory responses. Intuitively, if endothelial cells could be seeded on denuded vascular surfaces or on prostheses, the rate of thrombosis might be reduced.
73 1
Endothelium has a broad repertoire of clot modulating functions. It limits clot formation largely by deactivating thrombin via the antithrombin III and thrombomodulin systems. It suppresses platelet activation by elaborating prostacyclin and ecto-ATPase. Plasminogen activators and plasminogen activator inhibitors are products of endothelium. Therefore endothelium also exerts control over the dissolution of clot. During a response to injury, endothelium can produce clot promoting substances, such as platelet activating and thromboxane A, factor, tissue thromboplastin (TX&). Clotting factors V, von Willebrand’s factor, and plasminogen activator inhibitor are also made by unperturbed cells. It is widely believed that the patency of the grafts at the sites of vascular repair will be improved if endothelial clot promoting functions can be suppressed. Because endothelium inhibits and reverses clotting, endothelial linings inhibit experimental graft thromboses. Purely thrombotic occlusions tend to occur within a few weeks of grafting. Nevertheless, a very common event leading to later failure is neointimal fibrous hyperplasia. The development of hyperplasia begins with thrombin which sensitizes smooth muscle cells and fibroblasts to the effects of circulating growth factors. After complete endothelial resurfacing, the subjacent smooth muscle cells are no longer exposed to thrombin, nor are they directly exposed to serum-borne growth factors. Cellular growth factors are elaborated by endothelium, platelets, smooth muscle cells, and macrophages. Each of these cell types will remain in close contact with the vascular smooth muscle cells. Therefore rapid resurfacing of an injured vascular surface with endothelium is expected to reduce, but not eliminate, the hyperplastic response. Despite the theory and encouragement from laboratory animal implants, the reported clinical trials of endothelial seeding resulted in improvements in patency ranging from none to modest.“’ Is the theory wrong? The theory may be incomplete but not wrong. Three encouraging pieces of information have emerged from the clinical experience. Two pieces correlate with results in experimental animals: (1) adult human endothelium can be transplanted onto vascular prosthese?; (2) seeded human prostheses attract fewer platelets than unseeded ones.4 The third observation, although not particularly clear in animal models, is clinically glaring: (3) virtually all of the failures of seeded prostheses are caused by anastomotic neointimal fibrous hyperplasia.3 At the very root of the clinical results is the problem with massive inefficiency. Inefficient endothelial seeding means that fewer cells on the flow surface must replicate many more times to create an intact monolayer. The effect of prolonged exposure to thrombin and circulating growth factors on smooth muscle cells is obvious; however, replicating endothelium deposits its own growth factor until replication stops. Generally, replication stops when the monolayer of endothelium is complete.
732
Journal of VASCULAR SURGERY
Special communication
To analyze the inefficiency in creating a monolayer, endothelial seeding can be considered in three major steps: (1) removing endothelium from a donor vessel, (2) attaching endothelium to the recipient site, and (3) replication of the cells to confluence. In 1990 step 2 is 40% to 60% efficient, but steps 1 and 3 have considerable room for improvement. We harvested endothelium for seeding in 1976 with a steel wool pledget and mechanical abrasion. Approximately 10’ cells were obtained from each 1 cm’ of dog vein, from which lo” cells would be expected, an efficiency of 0.1%. Most harvested cells were dead. Viable cells were in large clusters and difficult to disperse, and there was a large contamination with subjacent smooth muscle cells. Burke1 et al.’ introduced enzymatic harvesting with bacterial collagenase and trypsin. The enzymatic methods reduced smooth muscle cell contamination, increased the viability of the harvested endothelium, improved the cell cluster size, and resulted in about 5 x LO4 cells/cm,’ an efficiency of 10%. Nevertheless, collagenase preparations are crude and vary substantially from lot to lot. Commercial efforts to standardize harvesting enzymes remain under evaluation. Pearce6 and Jarrell,’ working separately, applied enzymatic harvesting to microvascular endothelium. They achieved a tenfold increase in harvested endothelium but continued to have contamination with other cells. The second efficiency problem is that of cell retention on the graft. Rosenman et al.’ revealed that only 3.2% of the cells remained on polytetrafluoroethylene after 24 hours of perfusion.* Coupled with a 10% harvesting efficiency using enzymes, and a 60% adhesion efficiency, overall seeding efficiency is only about 0.19%. Cell retention is a difficult problem, one that is partly improved by use of hydrophilic substrates, adding cell attachment proteins (like fibronectin) and maturing the cell attachment sites before restoring blood flow. These measures are of limited value because cell loss is not simply a mechanical failure of attachment. Even the endothelium on autologous vein grafts is subject to massive loss. Studies of cell retention in neutropenic dogs revealed nearly complete cell retention during a 6-hour perfusion compared to 60% retention in untreated controls9 These results seem to incriminate the neutrophil in damaging the newly seeded endothelium during the early hours of flow. Messages pass between endothelium and neutrophils that trigger the attack. Some messengers are platelet activating factor, leukotriene B,, interleukin and thrombin,“’ but their respective roles in endothelial seeding remain unknown. However the endothelium is loosened, the loss of cells from the flow surface is the most important inefficiency of the seeding process. Because of the monolayer pattern of endothelium, few (no more than two-fold) “extra” cells can be supplied on a surface against flow-associated losses. Harvesting inefficiencies might be compensated by multiplying the yield in culture, but no more than 100% surface
coverage can be achieved. The question endothelial losses must be resolved.
of flow-associated
Endothelial cell seeding is the transplantation of vascular endothelial cells to denuded vascular surfaces. Seeding theoretically reduces the probability of graft or vessel thrombosis and of neointimal fibrous hyperplasia. Thus far, clinical seeding trials disclosed modest improvements in patency and the development of hyperplastic anastomotic lesions in failed grafts. Seeding inefficiency theoretically contributes to anastomotic hyperplasia. The inefficiency is linked to two steps in the seeding process, namely harvesting and cell retention. Of these, cell retention on the seeded surface is the more critical. Priorities for future research should be set first on the retention of seeded endothelium in vitro and second on improved and standardized methods of cell harvesting. Malcolm B. Herr& MLI Indiana Universiq School of Medicine St. T¢ Hospital Indianapolis, had.
REFERENCES 1. Faso1 R, Zilla P, Deutsch M, Grimm M, Fischlein T. Human endothelial cell seeding: Evaluation of its effectiveness by platelet parameters after one year. J VASC SURG 1989;9: 432-6. 2. Herring MB, Compton RS, LeGrand DR, Gardner AL, Madison DL, and Glover JL. Endothelial seeding of polytetrafluoroethylene popliteal bypasses. J VASC SURG 1987;6: 114-8. 3. Herring MB, LeGrand DR. The histology of seeded PTFE grafts in humans. Ann Vast Surg 1989;3:96-103. 4. Ortenwall I’, Wadenvik H, Kutti J, R&erg B. Endothelial cell seeding reduces thrombogenicity of Dacron grafts in humans. J VASC SURG 1990;11:403-10. 5. Graham LM, Burke1 WE, Ford JW, Vinter DW, Kahn RH, Stanley JC. Immediate seeding of enzymatically derived endothelium on Dacron vascular grafts; early experimental studies with autogenous canine cells. Arch Surg 1980;115: 1289-94. RB, Whitehill TA. Successml endo6. Pearce WH, Rutherford thelial seeding with omentally derived microvascular endothelial cells. J VAX SURG 1987;5:203-6, 1987. 7. Jarrell BE, Williams SK, Carabasi RA, Hubbard FA. lmmediate vascular graft monolayers using microvessel endothelial cells. In: Herring MB, Glover JL eds. Endothelial seeding in vascular surgery. Orlando: Grune & Stratton, Inc, 1987: 37-55. JE, Kemp&n&i RF, Pearce WH, Silberstein EB. 8. Rosenman Kinetics of endothelial cell seeding. J VAX SURC; 1985;77884. 9. Emerick S, Herring M, Arnold M, Baughman S, Reilly K, Glover J. Leukocyte depletion enhances cultured endothelial retention on vascular prostheses. J VAX SLJRG 1989;5: 342-7. 10. Tonnesen MG. Neutrophil-endothelial cell interaction: Mechanisms of neutrophil adherence to vascular endothelium. J Invest Dermatol 1989;93:53S-88.
Specialcommunication
level of a single focal adhesion, groups of adhesion plaques, a single cell, or groups of cells. It should be possible to manipulate the composition of the extracellular matrix to quantitate the effects on cell adhesion. It should also be possible to use competitive binding studies of adhesion proteins, as well as selective blocking of domains of adhesion proteins by use of specific antibodies, to obtain insights concerning the molecular mechanisms of cell adhesion. Focal adhesions in endothelial cells may represent a complex of proteins capable of signaling the effects of external forces such as shear stress to the interior of the cell. It would be of interest to investigate whether such structures and their associated proteins play a regulatory role in the responses of endothelial cells to external physical forces. Such experiments are critical to an understanding of the adhesion of endothelial cells on vascular grafts where cells must resist detachment by flow and where the presence of an endothelial lining is likely to assist graft patency and survival. Peteev F. Davies, PhD Pritekw School of Medicine The University of Chicgo Chicago, Ill.
REFERENCES 1. Holtz J, Forstermann U, Pohl V, Giesler M, Bassenge EJ. Flow-dependent, endothelium-mediated dilation of epicardial coronary arteries in conscious dogs: effects of cycle-oxygenase inhibition. J Cardiovasc Pharmacol 1984;6:1161-9. 2. Langille BL, O’Donnell F. Reductions in arterial diameter produced by chronic decreases in blood flow are endothelialdependent. Science 1986;231:405-8. 3. Davies PF. Endothelial cells, hemodynamic forces, and the localization of atherosclerosis. In: Ryan US., ed. Endothelial Cells, Vol. II. Boca Baton, Florida: CRC Press, 1988: 123-39. 4. Olesen SP, Clapham DE, Davies PF. Hemodynamic shear stressactivates a K + current in vascular endothelial cells. Nature 1988;331:168-170. 5. Frangos JA, Eskin SE, McIntyre LV, Ives CL Flow effects on prostacyclin production by cultured human endothelial cells. Science 1985;227:1477-9. Diamond SL, McIntyre LV, Share&in JB, Dieffenbach C, Scott KF, Eskin SG. J Cell Physiol (In press). Davies PF. How do vascular endothelial cells respond to flow? News in Physiol Sci 1989;4:22-6. Dull RO, Davies PF. Hemodynamic shear stress modulates signal-response coupling in vascular endothelial cells (In press). Paddock SW. Tandem scanning reflected-light microscopy of cell-substratum adhesions and stress fibres in Swiss 3T3 cells. J Cell Sci 1989;93: 142-6.
ENDOTHELIAL
CELL SEEDING
Endothelial cell seeding is the transplantation of vascular endothelial cells to denuded vascular surfaces. Endothelium is a crucial cell in the modulation of clotting and inflammatory responses. Intuitively, if endothelial cells could be seeded on denuded vascular surfaces or on prostheses, the rate of thrombosis might be reduced.
73 1
Endothelium has a broad repertoire of clot modulating functions. It limits clot formation largely by deactivating thrombin via the antithrombin III and thrombomodulin systems. It suppresses platelet activation by elaborating prostacyclin and ecto-ATPase. Plasminogen activators and plasminogen activator inhibitors are products of endothelium. Therefore endothelium also exerts control over the dissolution of clot. During a response to injury, endothelium can produce clot promoting substances, such as platelet activating and thromboxane A, factor, tissue thromboplastin (TX&). Clotting factors V, von Willebrand’s factor, and plasminogen activator inhibitor are also made by unperturbed cells. It is widely believed that the patency of the grafts at the sites of vascular repair will be improved if endothelial clot promoting functions can be suppressed. Because endothelium inhibits and reverses clotting, endothelial linings inhibit experimental graft thromboses. Purely thrombotic occlusions tend to occur within a few weeks of grafting. Nevertheless, a very common event leading to later failure is neointimal fibrous hyperplasia. The development of hyperplasia begins with thrombin which sensitizes smooth muscle cells and fibroblasts to the effects of circulating growth factors. After complete endothelial resurfacing, the subjacent smooth muscle cells are no longer exposed to thrombin, nor are they directly exposed to serum-borne growth factors. Cellular growth factors are elaborated by endothelium, platelets, smooth muscle cells, and macrophages. Each of these cell types will remain in close contact with the vascular smooth muscle cells. Therefore rapid resurfacing of an injured vascular surface with endothelium is expected to reduce, but not eliminate, the hyperplastic response. Despite the theory and encouragement from laboratory animal implants, the reported clinical trials of endothelial seeding resulted in improvements in patency ranging from none to modest.“’ Is the theory wrong? The theory may be incomplete but not wrong. Three encouraging pieces of information have emerged from the clinical experience. Two pieces correlate with results in experimental animals: (1) adult human endothelium can be transplanted onto vascular prosthese?; (2) seeded human prostheses attract fewer platelets than unseeded ones.4 The third observation, although not particularly clear in animal models, is clinically glaring: (3) virtually all of the failures of seeded prostheses are caused by anastomotic neointimal fibrous hyperplasia.3 At the very root of the clinical results is the problem with massive inefficiency. Inefficient endothelial seeding means that fewer cells on the flow surface must replicate many more times to create an intact monolayer. The effect of prolonged exposure to thrombin and circulating growth factors on smooth muscle cells is obvious; however, replicating endothelium deposits its own growth factor until replication stops. Generally, replication stops when the monolayer of endothelium is complete.
732
Journal of VASCULAR SURGERY
Special communication
To analyze the inefficiency in creating a monolayer, endothelial seeding can be considered in three major steps: (1) removing endothelium from a donor vessel, (2) attaching endothelium to the recipient site, and (3) replication of the cells to confluence. In 1990 step 2 is 40% to 60% efficient, but steps 1 and 3 have considerable room for improvement. We harvested endothelium for seeding in 1976 with a steel wool pledget and mechanical abrasion. Approximately 10’ cells were obtained from each 1 cm’ of dog vein, from which lo” cells would be expected, an efficiency of 0.1%. Most harvested cells were dead. Viable cells were in large clusters and difficult to disperse, and there was a large contamination with subjacent smooth muscle cells. Burke1 et al.’ introduced enzymatic harvesting with bacterial collagenase and trypsin. The enzymatic methods reduced smooth muscle cell contamination, increased the viability of the harvested endothelium, improved the cell cluster size, and resulted in about 5 x LO4 cells/cm,’ an efficiency of 10%. Nevertheless, collagenase preparations are crude and vary substantially from lot to lot. Commercial efforts to standardize harvesting enzymes remain under evaluation. Pearce6 and Jarrell,’ working separately, applied enzymatic harvesting to microvascular endothelium. They achieved a tenfold increase in harvested endothelium but continued to have contamination with other cells. The second efficiency problem is that of cell retention on the graft. Rosenman et al.’ revealed that only 3.2% of the cells remained on polytetrafluoroethylene after 24 hours of perfusion.* Coupled with a 10% harvesting efficiency using enzymes, and a 60% adhesion efficiency, overall seeding efficiency is only about 0.19%. Cell retention is a difficult problem, one that is partly improved by use of hydrophilic substrates, adding cell attachment proteins (like fibronectin) and maturing the cell attachment sites before restoring blood flow. These measures are of limited value because cell loss is not simply a mechanical failure of attachment. Even the endothelium on autologous vein grafts is subject to massive loss. Studies of cell retention in neutropenic dogs revealed nearly complete cell retention during a 6-hour perfusion compared to 60% retention in untreated controls9 These results seem to incriminate the neutrophil in damaging the newly seeded endothelium during the early hours of flow. Messages pass between endothelium and neutrophils that trigger the attack. Some messengers are platelet activating factor, leukotriene B,, interleukin and thrombin,“’ but their respective roles in endothelial seeding remain unknown. However the endothelium is loosened, the loss of cells from the flow surface is the most important inefficiency of the seeding process. Because of the monolayer pattern of endothelium, few (no more than two-fold) “extra” cells can be supplied on a surface against flow-associated losses. Harvesting inefficiencies might be compensated by multiplying the yield in culture, but no more than 100% surface
coverage can be achieved. The question endothelial losses must be resolved.
of flow-associated
Endothelial cell seeding is the transplantation of vascular endothelial cells to denuded vascular surfaces. Seeding theoretically reduces the probability of graft or vessel thrombosis and of neointimal fibrous hyperplasia. Thus far, clinical seeding trials disclosed modest improvements in patency and the development of hyperplastic anastomotic lesions in failed grafts. Seeding inefficiency theoretically contributes to anastomotic hyperplasia. The inefficiency is linked to two steps in the seeding process, namely harvesting and cell retention. Of these, cell retention on the seeded surface is the more critical. Priorities for future research should be set first on the retention of seeded endothelium in vitro and second on improved and standardized methods of cell harvesting. Malcolm B. Herr& MLI Indiana Universiq School of Medicine St. T¢ Hospital Indianapolis, had.
REFERENCES 1. Faso1 R, Zilla P, Deutsch M, Grimm M, Fischlein T. Human endothelial cell seeding: Evaluation of its effectiveness by platelet parameters after one year. J VASC SURG 1989;9: 432-6. 2. Herring MB, Compton RS, LeGrand DR, Gardner AL, Madison DL, and Glover JL. Endothelial seeding of polytetrafluoroethylene popliteal bypasses. J VASC SURG 1987;6: 114-8. 3. Herring MB, LeGrand DR. The histology of seeded PTFE grafts in humans. Ann Vast Surg 1989;3:96-103. 4. Ortenwall I’, Wadenvik H, Kutti J, R&erg B. Endothelial cell seeding reduces thrombogenicity of Dacron grafts in humans. J VASC SURG 1990;11:403-10. 5. Graham LM, Burke1 WE, Ford JW, Vinter DW, Kahn RH, Stanley JC. Immediate seeding of enzymatically derived endothelium on Dacron vascular grafts; early experimental studies with autogenous canine cells. Arch Surg 1980;115: 1289-94. RB, Whitehill TA. Successml endo6. Pearce WH, Rutherford thelial seeding with omentally derived microvascular endothelial cells. J VAX SURG 1987;5:203-6, 1987. 7. Jarrell BE, Williams SK, Carabasi RA, Hubbard FA. lmmediate vascular graft monolayers using microvessel endothelial cells. In: Herring MB, Glover JL eds. Endothelial seeding in vascular surgery. Orlando: Grune & Stratton, Inc, 1987: 37-55. JE, Kemp&n&i RF, Pearce WH, Silberstein EB. 8. Rosenman Kinetics of endothelial cell seeding. J VAX SURC; 1985;77884. 9. Emerick S, Herring M, Arnold M, Baughman S, Reilly K, Glover J. Leukocyte depletion enhances cultured endothelial retention on vascular prostheses. J VAX SLJRG 1989;5: 342-7. 10. Tonnesen MG. Neutrophil-endothelial cell interaction: Mechanisms of neutrophil adherence to vascular endothelium. J Invest Dermatol 1989;93:53S-88.
