Human trabecular cells. I. Establishment in tissue culture and growth characteristics

Transcription

1 Human trabecular cells I. Establishment in tissue culture and growth characteristics Jon R. Polansky, Robert N. Weinreb, John D. Baxter, and Jorge Alvarado After a careful dissection of trabecular tissue from recent postmortem specimens, human trabecular cells were established in tissue culture with 250 ng/ml fibroblast growth factor (FGF), Dulbecco's modified. Eagle's medium (DME), and 10% human serum. These conditions have allowed propagation of human trabecular cells for a number of passages at high density without apparent cellular degeneration. FGF increased the rate of cell division and the plating efficiency for trabecular cells but was not needed after cells had achieved, confiuency. Human trabecular cells had a pattern of growth which differed from human corneal keratocytes and human corneal endothelial cells compared at a similar passage. Propagation of human trabecular cells in vitro may provide a valuable source of experimental material to study the functional aspects of these cells which line the trabecular meshwork. Key words: trabecular meshwork, cell culture, fibroblast growth factor, glaucoma The trabecular meshwork of the eye is lined by cells of similar appearance in all layers of its structure. These cells have been referred to as endothelial cells or fibroblasts, but ultras tructu rally the cells of the trabecular meshwork have a similar appearance. 1 " 3 From the Departments of Ophthalmology and Medicine, the Howard Hughes Medical Institute and the Metabolic Research Unit, University of California, San Francisco. This work was supported by NIH grant EY02477, Fight for Sight postdoctoral grant F-301, and Damon Runyon and Walter Winchell Cancer Fund grant DRG- 117-F to Jon R. Polansky, Fight for Sight postdoctoral grant F-313 to Robert N. VVeinreb, and NIH grant EY02068 to Jorge Alvarado. John D. Baxter is an investigator of the Howard Hughes Medical Institute and supported this project with NIH grant EY Submitted for publication Sept. 27, Reprint requests: Jon R. Polansky, M.D., Endocrine Research Division 671 HSE, University of California, San Francisco, 3rd and Parnassus Aves., San Francisco, Calif These trabecular cells probably have an active functional role in the maintenance of the outflow pathway. In fact, it has been proposed that alterations in trabecular cell synthesis of connective tissue elements 4 " 7 or phagocytosis 8 " 10 may contribute to the pathogenesis of primary open-angle glaucoma and some obstructive glaucomas. Studies of trabecular cell function using animal systems in vivo have been difficult, and it is possible that human trabecular cells may have different properties from cells which line the outflow pathway in animals. To study the structural and functional activity of human trabecular cells, it was desirable to establish these cells in tissue culture. Previously, Schachtschabel et al. 4 have shown that monkey trabecular cells could be grown for a number of generations, but the propagation of human trabecular cells in vitro has remained a problem. Here we report conditions which allow /79/ $00.70/ Assoc. for Res. in Vis. and Ophthal., Inc. 1043

2 1044 Polansky et al. Invest. Ophthalmol. Visual Sci. October 1979 Fig. 1. Propagation of cells from trabecular explant. Specimens were maintained in culture with FGF according to Methods. Fig. 1A. Typical dissected trabecular specimen (montage of meridional section). The sharp border at the right margin represents a cut made with a blade breaker during the dissection of trabecular tissue from scleral spur. the establishment of human trabecular cells in tissue culture and describe the growth characteristics of these cells. Trabecular cells are a distinct cell type when compared to corneal keratocytes and corneal endothelium. In a paper that will follow this one, 3 we report that the distinguishing ultrastructural features of human trabecular cells are preserved in culture and that human trabecular cells in vitro demonstrate a wide range of functional capabilities. Methods Three human cell lines were established from trabecular specimens obtained postmortem from patients 18, 52, and 65 years old with no personal or family history of glaucoma. Within the first 12 hr following death, trabecular tissue was dissected under a Zeiss operating microscope. Explants of dissected trabecular tissue were placed in 10% human serum with Dulbecco's modified Eagle's medium (DME), 2 mm glutamine, 50 /xg/ml gentamicin, and 2.5 /xg/ml Fungizone and incubated at 37 under a 10% CO 2 atmosphere. No proteolytic agents or tissue dissociation techniques were employed, since possible alterations in trabecular cells might occur due to these methods. If tissue explants were left undisturbed under a sterile screen for 1 to 2 weeks, they attached to the culture dish. By the third week in culture, cells migrated spontaneously from all areas of the explant. This occurred slightly sooner when fibroblast growth factor (FGF) (250 ng/ml) was added to the cultures but also occurred with no FGF added. A larger cell mass was achieved with FGF, but cultures otherwise appeared identical. After 1 week of plating, the tissue explant was removed, and the cells were passaged. All cultures were stimulated with FGF every other day with media changes every fourth day. Also, multiple platings were performed on each original tissue specimen by transferring them to new tissue culture dishes. Confluent cultures were passaged at a ratio of 1:4 or 1:6 with the use of 0.1% trypsin (Difco). Karyotyping of trabecular cultures after four passages showed that the diploid character of cells (46n) was preserved and that no gross chromosomal aberration occurred due to FGF or serial propagation in culture. To compare the growth properties of neighboring corneal cells to trabecular cells in vitro, human corneal endothelial cells and keratocytes were obtained from a 65-year-old patient and propagated under conditions identical to those described above, with FGF used to stimulate cell division. However, human corneal endothelium grew very poorly and could not be subcultured beyond the second passage. Results To establish human trabecular cells in culture, it was important to achieve a careful dissection of the meshwork from surrounding tissues. Once cells were propagated from these dissected specimens, it was important to demonstrate that all platings from each specimen yielded the same cell type. In addition, since human corneal keratocytes and human corneal endothelium represented potential contaminant cells, the growth proper-

3 Volume 18 Number 10 Human trabecular cells in tissue culture 1045 Fig. 1, B to D. B, Primary growth of trabecular cells in culture. Cells are actively proliferating from all areas of the explant, which is seen as a dark triangular shape. C, Growing trabecular cells at the periphery of the primary culture; arrows indicate migrating trabecular cells with "ruffled" edges. D, Passaged trabecular cells approaching confluency. ties of these cells were examined and compared to those of trabecular cells. A typical cross-section of a dissected trabecular specimen is shown in Fig. 1A. Specimens used for explant showed many trabecular beams and the absence of scleral spur, ciliary muscle, cornea, iris, and Schlemm's canal structures. Multiple cross-sections showed that a clean dissection of trabecular tissue had been achieved. When these sections were evaluated with electron microscopy, the vast majority of cells were a homogeneous population of trabecular lining cells. No cell types from contiguous structures were seen, although rarely a Schlemm's canal cell or a macrophage could be detected in some specimens. The trabecular tissue was placed in culture according to Methods. The initial growth of human trabecular cells from the explant is shown in Fig. 1, B and C. Active proliferation of trabecular cells was observed from all areas of the tissue specimen. Proliferating cells migrated from the tissue specimens, forming an aggregate cell mass surrounding the explant. Often the migrating trabecular cells showed ruffled borders as they detached from the main cell mass (Fig. 1, C). Occasionally migrating cells formed "satellite" colonies of cells with identical morphology to the main cell mass. After 1 week of plating, trabecular cultures were passaged according to Methods. The passaged cells appeared homogeneous, and different cell types were not observed. Fig. 1, D shows second-passage trabecular cells as they first became confluent; cells from earlier and later passages appeared identical to these. Although FGF was used for the passage of

4 1046 Polansky et al. Invest. Ophtluihnol. Visual Sci. October i 4 6 Days I i 8 + FGF nofgf i 10 Fig. 2. Growth response for human trabecular cells in vitro; effect of FGF. Third-passage trabecular cells from a 65-year-old patient were plated at 2 X 10 :! cells/cm 2 for 8 hr. Cell counts were performed in duplicate after plating on every other day. all cultures, the mitogenic effect observed with this factor was greater in some trabecular lines than in others. Cells from older patients appeared to be more dependent on FGF than cells from younger patients. Fig. 2 shows a quantitative growth experiment on human trabecular cells from a 68-year-old patient. Third-passage human trabecular cells were plated at a density of 2 X 10 3 cells/cm 2. These had a doubling time of approximately 3 days in the presence of FGF; cells grown without FGF showed a much longer doubling time. At the end of 10 days in culture, the trabecular cells with growth factor had achieved a density of 10.8 X 10 3 cells/cm 2 (almost confluent), whereas cells grown without FGF had a density of 3.6 X 10 3 cells/ cm 2. Cells from this patient maintained without FGF were more difficult to passage, and their plating efficiency fell from the 75% to 85% usually observed to approximately 35%. When cells had achieved confluent density, FGF was no longer required; however, addition of FGF showed that trabecular cells were not strictly contact-inhibited, and cells and cell borders were seen to overlap. In examining the effect of subculture on human trabecular cells, it was of interest to compare the morphological appearance of early-passage cells to that of later passages. During the growing phase and early confluency, human trabecular cells appeared similar in all passages. If trabecular cells were maintained at confluency for 1 to 2 weeks, certain differences between early and late passages were seen. Fig. 3, A, B and C, shows human trabecular cells maintained at confluency after the first, second, and fourth passages, respectively. Cells from the first passage showed a rather orderly array, especially in certain areas where the cells were closely packed. Cells from the second passage had fewer areas in which the cells packed tightly and showed a less orderly arrangement. In later passages (the typical appearance is seen in Fig. 3, C), more random cell overlapping was observed, but the viability of the individual cells was preserved. A slight increase in the number of degenerating cells was observed after the fifth passage in culture. To further evaluate the growth pattern of human trabecular cells, these were compared with neighboring corneal cells (human corneal keratocytes and corneal endothelium). All cells were grown under identical conditions and compared at the second passage. Cells were plated at 5 X 10 3 cells/cm 2 and maintained for 2 weeks according to Methods, with FGF. Human trabecular cells approached confluency between days 4 and 5 in culture and after 2 weeks (Fig. 4, A) were seen to grow as a single cell layer, with occasional areas in which cells and cell borders overlapped. Human corneal endothelium (Fig. 4, B) formed contact-inhibited sheets of cells in vitro and had much poorer growth properties than did trabecular cells or corneal keratocytes. Corneal endothelium formed an incomplete monolayer of very large, polygonal, cells which could not be effectively grown after the second passage. Human corneal keratocytes (Fig. 4, C) proliferated more rapidly than either of the other two cell types and readily overgrew one another in multiple layers, with an appearance similar to that

5 Volume 18 Number.10 Human trabecular cells in tissue culture 1047 "ffll Fig. 3. Appearance of confluent cultures ol human trabecular cells. A, First passage. B, Second passage. C, Fourth passage. Cells were maintained for 1 to 2 weeks after confluency according to Methods. seen in cultures of skin fibroblasts. As a further comparison, keratocyte and trabecular cultures were examined after cultures were fixed, embedded, and sectioned. Fig. 4, D and E, show these preparations at approximately 4 times higher magnification than that shown in Fig. 4, A to C. When compared to keratocytes, trabecular cells appeared wider and had more numerous cell extensions and processes. Keratocytes were cylindrically shaped and showed only rare cellular processes. Therefore human trabecular cells had different growth properties from human corneal endothelium or human corneal keratocytes, and neither of these neighboring cells was likely to be a contaminant in our trabecular cultures. Discussion These studies demonstrate that human trabecular cells can be established in tissue culture and effectively passaged. Use of recent postmortem specimens has provided a readily available source of experimental material for tissue culture. A careful dissection of this tissue was important to make sure that only trabecular meshwork was obtained, and this helped to assure that the cells grown in tissue culture were not contaminant cell types. Under the culture conditions employed, FGF aided in the establishment of the trabecular cultures and stimulated cell division. Previously, it has been shown that FGF is effective in stimulating cell division for vascular endothelium and corneal endothelium in vitro 11 ; in addition, FGF appears to be a "survival" factor for certain cell types. 12 FGF appeared to prevent cellular degeneration in some of our human trabecular lines, especially those obtained from older patients. Use of fetal calf serum instead of human serum

6 1048 Polansktj et al. Invest. Ophthahnol. Visual Sci, October 1979 Fig. 4. Comparison of human trabecular cells to human corneal cells, in vitro at the second passage after 2 weeks in tissue culture according to Methods. A, Trabecular cells. B, Corneal endothelial cells. C, Corneal keratoeytes. (A to C were photographed under phase contrast at identical magnifications.) D, Trabecular cells. E, Corneal keratocytes, after sectioning in the flat and staining with 1% toluene blue. (Magnification was approximately 4x that shown in A to C). may reduce the need for mitogenic stimulation by FGF, but fetal calf serum contains a variety of factors which may complicate hormonal studies on the trabecular cells. Recently we have found that treating tissue culture surfaces with 0.2% gelatin improves with plating efficiency and growth characteristics of the different trabecular lines. Future improvements in cell culture techniques will examine other mitogenic factors and different substrates for cell growth. When we examined the growth properties of human trabecular cells in culture, the trabecular cell appeared as a distinct cell type in comparison with neighboring corneal cell types of the eye. We have also examined 3 the ultrastructural features of trabecular cells grown in vitro through four passages and have found that the distinctive ultrastructural characteristics of trabecular cells seen in vivo are preserved in our cultured specimens. When the appearance of confluent trabecular cells was compared at different passages, it was noted that later-passage cells did not pack in as orderly as did those of the first passage. This may be related to a decrease in

7 Volume 18 Number 10 Human trabecular cells in tissue culture 1049 collagen production which is commonly seen in the passage of primary mammalian lines. Nevertheless, the trabecular cells obtained from the different patients could be passaged at least five times without a noticeable increase in cellular degeneration or decrease in proliferative capacity. Therefore the propagation of human trabecular cells should allow a variety of studies to be performed on these cells which are not possible in vivo. In particular, biochemical, morphological, and hormonal studies may be readily performed with trabecular cells grown in tissue culture, and experimental conditions can be reproducibly controlled. In addition, it is possible that studies of human trabecular cells in culture will allow the testing of current theories of the pathogenesis of glaucoma and may allow the evaluation of new pharmacological agents. We thank Dr. Gospodarowicz who has provided us with valuable advice on endothelial cell biology and fibroblast growth factor, and Dr. Steven Kramer for helpful discussions. REFERENCES 1. Hogan, M.J., Alvarado, J.A., and Weddell, J.E.: The limbus. In: Histology of the Human Eye, Philadelphia, 1977, VV.B. Saunders Co., pp Vegge, T.: Ultrastructure of normal human trabecular endothelium, Acta Ophthalmol 41:193, Alvarado, J.A., Wood, I., and Polansky, J.R.: Human trabecular cells. II. Ultrastructural features and biological activity in vitro, (to be submitted to INVEST. OPHTHALMOL. VISUAL SCI.). 4. Schachtschabel, D.O., Bigalke, B., and Rohen, J.W.: Production of glycosaminoglycans by cell cultures of the trabecular meshwork of the primate eye, Exp. Eye. Res. 24:71, Polansky, J.R., Gospodarowicz, D., Weinreb, R., and Alvarado, J. Human trabecular meshwork cell culture and glycosaminoglycan synthesis, INVEST. OPHTHALMOL. VISUAL SCI. 17(ARVO SuppI.):207, Rohen, J.W., and Schachtschabel, D.O.: Morphologic and biochemical studies of human trabecular meshwork in tissue culture, INVEST. OPHTHALMOL. VISUAL SCI. 17(ARVO Suppl.):207, Francois, J.: Corticosteroid glaucoma, Metabolic Ophthalmol. 2:3, Grierson, I., and Lee, W.R.: Erythrocyte phagocytosis in the human trabecular meshwork, Br. J. Ophthalmol. 57:400, Rohen, J. W., and Vanderzypen, E.: The phagocytic activity of the trabecular meshwork endothelium, Albrecht Von Graefes Arch. Klin. Exp. Ophthalmol. 175:143, Bill, A.: The draingage of aqueous humor, INVEST. OPHTHALMOL. 14:1, Gospodarowicz, D., Greenburg, G., Bialecki, H., and Zetter, B.R.: Factors involved in the modulation of cell proliferation in vivo and in vitro: the role of fibroblast and epidermal growth factors in the proliferative response of mammalian cells, In Vitro 14:85, Gospodarowicz, D., Moran, J.S., Braun, D.L., and Birdvvell, C.R.: Clonal growth of bovine vascular endothelial cells: fibroblast growth factor as a survival agent, Proc. Natl. Acad. Sci., U.S.A., 73:4120, 1976.