LETTERS TO THE EDITORS The Editors invite readers to submit letters commenting on the contents of articles that appear in the Journal. Also welcome are brief communications in letter form reporting investigative or clinical observations without extensive documentation and with brief bibliography (five titles or less), not requiring peer review but open to critique by readers. Letters to the Editors should be no more than 500 words in length and they may have to be edited for publication.
Cells derived from omental fat tissue and used for seeding vascular prostheses are not endothelial in origin To the Editors:
We read with interest the recent article by Visser et al. concerning the nature of cells derived from human omentum. We too have been working with this source of cells, and although we agree that further characterization is required, we do not feel it yet proved that these cells are mesothelial in origin. Although Visser et al. have studied a wide range of features of their cells, they have not made a direct comparison with mesothelial cells. It is not enough to compare these cells with human umbilical vein endothelial cells since abundant evidence exists that endothelium from different sites can differ markedly from macrovascular endothelium, such as that derived from the human umbilical vein. 2-4The best described example of this is that microvascular endothelium lacks the Weibel-Palade bodies (thought to be the storage granules for factor-VIii) of rnacrovascnlar endothelium and does not therefore stain in a granular fashion for von Willebrand factor, but rather in the diffuse manner the authors describe. We have looked at the staining of human omental microvascular endothelial cells (HOTMECs) and compared it with staining patterns seen in human mesothelium, fibroblasts, and umbilical vein endothelial cells, using both "endothelial specific" markers, namely polyclonal anti-factor VIII, EN4, PAL-E, and QBEND-40, and monoclonals against Thy-1 and CDw44 glycoprotein, which are not thought to target endothelium but which are found on cells of connective tissue origin. 2-7 When cytospins from the original isolates (that is, before tissue culture but after percoll density gradient centrifugation) were examined with use of immunofluorescence microscopy, only occasional cells (< 1%) were found to be positive for factor VIII related antigen. Cells positive for Thy-1 and CDw44 glycoprotein comprised 15% to 35% of the initial isolate, and positive cells were seen both as single cells and in combination with negative cells in cell clumps. Cells positive for EN4, PAL-E, and QBEND-40 were not seen. By first passage in tissue culture, however, the appearances were much more homogeneous. Most cells were now positive for factor-VIii related antigen, Thy- 1 and CDw44 glycoprotein. A small proportion ( < 5%) were also positive for the endothelial markers EN4, PAL-E, and QBEND-40. Fresh mesothelial smears were positive for Thy-1 and CDw44 glycoprotein but negative for factor VIII related antigen and other "endothelial" markers. Cultured human
skin fibroblasts were positive for Thy-1 and CDw44 glycoprotein but negative for all the other markers studied including factor VIII related antigen. Human umbilical vein endothelial cells in cxflture and at initial harvest were positive for factor VIII related antigen, EN4, PAL-E, and QBEND-40 but negative for Thy-1 and CDw44 glycoprotein. O f particxflar interest is that Thy-1 and CDw44 glycoprotein positive cells were present in cell clumps together with negative cells. This would suggest that these were not clumps of mesothelium, which would be expected to demonstrate uniform staining. In addition we were unable to demonstrate that mesothelium stained positive for yon Willebrand factor, which oaltured HOTMECs clearly do. It is possible that these cells are pericytic in nature as pericytes are known to outgrow endothelial cells in culture. 8'9 In conclusion, therefore, although most of our findings agree with those ofVisser et al., we feel that the identity of these cells remains unproved. They do produce prostacyclln in similar quantities and by similar mechanisms to macrovascular endothelium and they do produce large amounts of tPA. 1°,11 We feel that this source of cells is worthy of further study and may yet prove to be the best source of cells for endothelial ceil seeding trials in humans. G. Stansby N. Shukla B. Fuller G. Hamilton
The Academic Department of Surgery The Royal Free Hospital Pond St. Hampstead, London NW3 3QG United Kingdom 24/41/32347 REFERENCES
1. Visser MJT, van Bockel JH, van Muijen GNP, van Hinsbergh VWM. Cells derived from omental fat tissue and used for seeding vascular prostheses are not endothelial in origin. J VASCSURG1991;13:373-81. 2. Kumar S, West DC, Ager A. Heterogeneity in endothelial cells from large vessels and microvessels. Differentiation 1987;36: 57-70. 3. Wagner WH, Henderson RM, Hicks HE, Banes AJ, Johnson G. Differences in morphology, growth rate and protein synthesis between cultured arterial and venous endothelial cells. J VASCSURG1988;8:19203-519. 4. Pearson JD. Isolation and culture of microvascxflarendothelium. In: Warren JB, ed. The endothelium: an introduction to current research. New York: Wiley-Liss, 1990: 295-305. 583
584
Journal of VASCULAR SURGERY
Letters to the Editors
5. McKenzie JL, Fabre JW. Human Thy- 1: Unusual localization and possible functional significance in lymphoid tissues. J Immunol 1981;126:843-50. 6. Dalchau R, Daar AS, Fabre JW. The Thy-1 molecule in man. In: ReifA, Schlensinger M, ed. Thy-l: Immunology, neurology and therapeutic applications. New York: Marcel Dekker, 1989:185-96. 7. Flanagan BF, Dalchau R, Allen AK, Daar AS, Fabre JW. Chemical composition and tissue distribution of the human CDw44 glycoproteiu. Immunology 1989;67:167-75. 8. Orlidge A, D'Amore PA. Inhibition of capillary endothelial cell growth by pericytes and smooth muscle cells. J Cell Biol 1987;105:1455-62. 9. Zetter BR. Culture of capillary endothelial cells. In: Jaffe EA, ed. Biology of endothelial cells, Boston: Martinus Nijhoff, 1984:14-26. 10. Stansby G, Shukla N, Fuller B, Hamilton G. Characterisation of microvascular endothelial cells derived from human fat. Biochem Soc Transactions 1991;19:168S. 11. Stansby G, Shukla N, Hamilton G, Jeremy J. Comparison of prostanoid synthesis in cultured human vascular endothelial cells derived from omentum and umbilical vein. Eur J" Vase Surg 1991;5:501-6.
Reply To the Editors:
We appreciate the comments made by Dr. Stansby and his colleagues with respect to the evaluation of the origin of cells derived from omental fat tissue. Their main criticism is that we did not compare characteristics of omental fat derived cells with those of mesothelial cells derived from smears. However, we cannot agree with their statement that we did not prove that these cells are of mesothelial nature. We and others have previously studied the characteristics of mesothelial cells in histologic sections, 12 using monoclonal antibodies directed against intermediate filaments of the cytoskeleton. Therefore it is justified to state that the combined presence of these intermediate filaments is a marker for the mesothelial nature o f the cultured cells we described. The finding that these cells are mesothelial cells was further strengthened by their abundant expression o f surface microvilli, a cellular characteristic that has been demonstrated as a quality of mesothelial cells in histologic sections as well as in culture. 3-4 The statement that microvascular endothelial cells lack Weibel-Palade bodies is probably not true. Human foreskin microvascular endothelial cells produce a large number of granules containing yon Willebrand factor? In addition, we have isolated microvessels from human subcutaneous fat. The endothelial cells from these microvessels also display a granular staining of von Willebrand factor antigen. 6 In the latter paper we have indicated that a small number o f endothelial cells are sometimes encountered between the abundant mesothelial cells. It is therefore of importance that we have described in the article that Dr. Stansby debates that not only the large majority of the cells are mesothelial in nature, but that endothelial cells are
rapidly displaced by mesothelial cells when they are put together in a culture dish. We have indicated that this results in a pure culture of mesothelial cells. Subsequently we have advocated the use of various markers, in particular cytokeratins to establish the nature of the cells. Dr. Stansby describes the phenomenon that during cell culture a higher proportion of cells derived from omental fat tissue stain positively for the endothelial cell markers EN-4, PAL-E, and QBEND-40 when compared to cells in the original isolate. This suggests the outgrowth of endothelial cells, which is in contrast with our findings 7 as well as with the findings that Dr. Stansby reported previously.8 Moreover, Dr. Stansby describes the presence of clumps of cells that stain positively for Thy-1 and CDw44 glycoproteins together with negative cells, suggesting a mixture o f cells. As stated above, we sometimes find a small proportion of endothelial cells between the mesothelial cells, but in contrast to the findings of Dr. Stansby these endothelial cells are rapidly replaced by the mesothelial cells in culture. We fully agree with Dr. Stansby that omental fat tissue as source for seeding purposes is worthy of further study. Mesothelial cells indeed produce large amounts of tissuetype plasminogen activator (tI'A) as we have previously demonstrated. 6 However, recently a characteristic o f these cells that would not be beneficiary to cell seeding was also demonstrated: the abundant expression of tissue factor by both native and cultured cells derived from omental tissue. 9
In conclusion, although the processing o f omental fat tissue with techniques described in the literature yields predominantly mesothelial cells, omental fat remains a possible and interesting source of cells to seed vascular prostheses to improve their fimction. Michel J. T. Visser, 214D f. Hajo van Bockel, 3liD Victor W. 34. van Hinsbe~h, PhD Goos N. P. van 34uijen, PhD
Department of Vascular Surgery University Hospital Leiden P.O. Box 9600 2300 RC Leiden The Netherlands 24/41/32348 REFERENCES
1. Van Muijen GNP, Ruiter D~ Warnaar SO.. Coexpression of intermediate filament polypeptides in human fetal and adult tissues. Lab Invest 1987;57:359-69. 2. La Rocca P, Rheinwald IG. Coexpression of simple epithelial keratins and vimentin by human mesothelium and mesothelioma in vivo and in culture. Cancer Res 1984;44:2991-9. 3. Takahashi K, Goto T, Mukai K, Sawasaki Y, Hata I. Cobblestone monolayer cells from human omental adipose tissue are possibly mesothelial, not endothelial. In Vitro Cell Develop Biol 1989;25:109-1I. 4. Hessendahl H, Larsen JF. Ultrastrucmre of human yolk sac: endoderm, mesenchyme, tubules and mesothelium. Am J Anat 1969;I26:315-36.
Cells derived from omental fat tissue and used for seeding vascular prostheses are not endothelial in origin To the Editors:
We read with interest the recent article by Visser et al. concerning the nature of cells derived from human omentum. We too have been working with this source of cells, and although we agree that further characterization is required, we do not feel it yet proved that these cells are mesothelial in origin. Although Visser et al. have studied a wide range of features of their cells, they have not made a direct comparison with mesothelial cells. It is not enough to compare these cells with human umbilical vein endothelial cells since abundant evidence exists that endothelium from different sites can differ markedly from macrovascular endothelium, such as that derived from the human umbilical vein. 2-4The best described example of this is that microvascular endothelium lacks the Weibel-Palade bodies (thought to be the storage granules for factor-VIii) of rnacrovascnlar endothelium and does not therefore stain in a granular fashion for von Willebrand factor, but rather in the diffuse manner the authors describe. We have looked at the staining of human omental microvascular endothelial cells (HOTMECs) and compared it with staining patterns seen in human mesothelium, fibroblasts, and umbilical vein endothelial cells, using both "endothelial specific" markers, namely polyclonal anti-factor VIII, EN4, PAL-E, and QBEND-40, and monoclonals against Thy-1 and CDw44 glycoprotein, which are not thought to target endothelium but which are found on cells of connective tissue origin. 2-7 When cytospins from the original isolates (that is, before tissue culture but after percoll density gradient centrifugation) were examined with use of immunofluorescence microscopy, only occasional cells (< 1%) were found to be positive for factor VIII related antigen. Cells positive for Thy-1 and CDw44 glycoprotein comprised 15% to 35% of the initial isolate, and positive cells were seen both as single cells and in combination with negative cells in cell clumps. Cells positive for EN4, PAL-E, and QBEND-40 were not seen. By first passage in tissue culture, however, the appearances were much more homogeneous. Most cells were now positive for factor-VIii related antigen, Thy- 1 and CDw44 glycoprotein. A small proportion ( < 5%) were also positive for the endothelial markers EN4, PAL-E, and QBEND-40. Fresh mesothelial smears were positive for Thy-1 and CDw44 glycoprotein but negative for factor VIII related antigen and other "endothelial" markers. Cultured human
skin fibroblasts were positive for Thy-1 and CDw44 glycoprotein but negative for all the other markers studied including factor VIII related antigen. Human umbilical vein endothelial cells in cxflture and at initial harvest were positive for factor VIII related antigen, EN4, PAL-E, and QBEND-40 but negative for Thy-1 and CDw44 glycoprotein. O f particxflar interest is that Thy-1 and CDw44 glycoprotein positive cells were present in cell clumps together with negative cells. This would suggest that these were not clumps of mesothelium, which would be expected to demonstrate uniform staining. In addition we were unable to demonstrate that mesothelium stained positive for yon Willebrand factor, which oaltured HOTMECs clearly do. It is possible that these cells are pericytic in nature as pericytes are known to outgrow endothelial cells in culture. 8'9 In conclusion, therefore, although most of our findings agree with those ofVisser et al., we feel that the identity of these cells remains unproved. They do produce prostacyclln in similar quantities and by similar mechanisms to macrovascular endothelium and they do produce large amounts of tPA. 1°,11 We feel that this source of cells is worthy of further study and may yet prove to be the best source of cells for endothelial ceil seeding trials in humans. G. Stansby N. Shukla B. Fuller G. Hamilton
The Academic Department of Surgery The Royal Free Hospital Pond St. Hampstead, London NW3 3QG United Kingdom 24/41/32347 REFERENCES
1. Visser MJT, van Bockel JH, van Muijen GNP, van Hinsbergh VWM. Cells derived from omental fat tissue and used for seeding vascular prostheses are not endothelial in origin. J VASCSURG1991;13:373-81. 2. Kumar S, West DC, Ager A. Heterogeneity in endothelial cells from large vessels and microvessels. Differentiation 1987;36: 57-70. 3. Wagner WH, Henderson RM, Hicks HE, Banes AJ, Johnson G. Differences in morphology, growth rate and protein synthesis between cultured arterial and venous endothelial cells. J VASCSURG1988;8:19203-519. 4. Pearson JD. Isolation and culture of microvascxflarendothelium. In: Warren JB, ed. The endothelium: an introduction to current research. New York: Wiley-Liss, 1990: 295-305. 583
584
Journal of VASCULAR SURGERY
Letters to the Editors
5. McKenzie JL, Fabre JW. Human Thy- 1: Unusual localization and possible functional significance in lymphoid tissues. J Immunol 1981;126:843-50. 6. Dalchau R, Daar AS, Fabre JW. The Thy-1 molecule in man. In: ReifA, Schlensinger M, ed. Thy-l: Immunology, neurology and therapeutic applications. New York: Marcel Dekker, 1989:185-96. 7. Flanagan BF, Dalchau R, Allen AK, Daar AS, Fabre JW. Chemical composition and tissue distribution of the human CDw44 glycoproteiu. Immunology 1989;67:167-75. 8. Orlidge A, D'Amore PA. Inhibition of capillary endothelial cell growth by pericytes and smooth muscle cells. J Cell Biol 1987;105:1455-62. 9. Zetter BR. Culture of capillary endothelial cells. In: Jaffe EA, ed. Biology of endothelial cells, Boston: Martinus Nijhoff, 1984:14-26. 10. Stansby G, Shukla N, Fuller B, Hamilton G. Characterisation of microvascular endothelial cells derived from human fat. Biochem Soc Transactions 1991;19:168S. 11. Stansby G, Shukla N, Hamilton G, Jeremy J. Comparison of prostanoid synthesis in cultured human vascular endothelial cells derived from omentum and umbilical vein. Eur J" Vase Surg 1991;5:501-6.
Reply To the Editors:
We appreciate the comments made by Dr. Stansby and his colleagues with respect to the evaluation of the origin of cells derived from omental fat tissue. Their main criticism is that we did not compare characteristics of omental fat derived cells with those of mesothelial cells derived from smears. However, we cannot agree with their statement that we did not prove that these cells are of mesothelial nature. We and others have previously studied the characteristics of mesothelial cells in histologic sections, 12 using monoclonal antibodies directed against intermediate filaments of the cytoskeleton. Therefore it is justified to state that the combined presence of these intermediate filaments is a marker for the mesothelial nature o f the cultured cells we described. The finding that these cells are mesothelial cells was further strengthened by their abundant expression o f surface microvilli, a cellular characteristic that has been demonstrated as a quality of mesothelial cells in histologic sections as well as in culture. 3-4 The statement that microvascular endothelial cells lack Weibel-Palade bodies is probably not true. Human foreskin microvascular endothelial cells produce a large number of granules containing yon Willebrand factor? In addition, we have isolated microvessels from human subcutaneous fat. The endothelial cells from these microvessels also display a granular staining of von Willebrand factor antigen. 6 In the latter paper we have indicated that a small number o f endothelial cells are sometimes encountered between the abundant mesothelial cells. It is therefore of importance that we have described in the article that Dr. Stansby debates that not only the large majority of the cells are mesothelial in nature, but that endothelial cells are
rapidly displaced by mesothelial cells when they are put together in a culture dish. We have indicated that this results in a pure culture of mesothelial cells. Subsequently we have advocated the use of various markers, in particular cytokeratins to establish the nature of the cells. Dr. Stansby describes the phenomenon that during cell culture a higher proportion of cells derived from omental fat tissue stain positively for the endothelial cell markers EN-4, PAL-E, and QBEND-40 when compared to cells in the original isolate. This suggests the outgrowth of endothelial cells, which is in contrast with our findings 7 as well as with the findings that Dr. Stansby reported previously.8 Moreover, Dr. Stansby describes the presence of clumps of cells that stain positively for Thy-1 and CDw44 glycoproteins together with negative cells, suggesting a mixture o f cells. As stated above, we sometimes find a small proportion of endothelial cells between the mesothelial cells, but in contrast to the findings of Dr. Stansby these endothelial cells are rapidly replaced by the mesothelial cells in culture. We fully agree with Dr. Stansby that omental fat tissue as source for seeding purposes is worthy of further study. Mesothelial cells indeed produce large amounts of tissuetype plasminogen activator (tI'A) as we have previously demonstrated. 6 However, recently a characteristic o f these cells that would not be beneficiary to cell seeding was also demonstrated: the abundant expression of tissue factor by both native and cultured cells derived from omental tissue. 9
In conclusion, although the processing o f omental fat tissue with techniques described in the literature yields predominantly mesothelial cells, omental fat remains a possible and interesting source of cells to seed vascular prostheses to improve their fimction. Michel J. T. Visser, 214D f. Hajo van Bockel, 3liD Victor W. 34. van Hinsbe~h, PhD Goos N. P. van 34uijen, PhD
Department of Vascular Surgery University Hospital Leiden P.O. Box 9600 2300 RC Leiden The Netherlands 24/41/32348 REFERENCES
1. Van Muijen GNP, Ruiter D~ Warnaar SO.. Coexpression of intermediate filament polypeptides in human fetal and adult tissues. Lab Invest 1987;57:359-69. 2. La Rocca P, Rheinwald IG. Coexpression of simple epithelial keratins and vimentin by human mesothelium and mesothelioma in vivo and in culture. Cancer Res 1984;44:2991-9. 3. Takahashi K, Goto T, Mukai K, Sawasaki Y, Hata I. Cobblestone monolayer cells from human omental adipose tissue are possibly mesothelial, not endothelial. In Vitro Cell Develop Biol 1989;25:109-1I. 4. Hessendahl H, Larsen JF. Ultrastrucmre of human yolk sac: endoderm, mesenchyme, tubules and mesothelium. Am J Anat 1969;I26:315-36.
