Vitreous Modulation of Migration and Proliferation of Retinal ... - IOVS

Investigative Ophthalmology & Visual Science, Vol. 30, No. 9, September 1989 Copyright © Association for Research in Vision and Ophthalmology

Vitreous Modulation of Migration and Proliferation of Retinal Pigment Epithelial Cells In Vitro Bernd Kirchhof,* Erika Kirchhof,* Stephen J. Ryan, and Nino Sorgente Retinal pigment epithelial (RPE) cell migration and proliferation are believed to play a role in the pathogenesis of proliferative vitreoretinopathy (PVR). Since PVR develops in situations where vitreous contacts the RPE, we sought to determine whether human vitreous contains factors that stimulate proliferation and migration of RPE cells. We found that postmortem human vitreous stimulates migration but not proliferation of human RPE cells under serum-free conditions in vitro. Stimulation of proliferation of RPE cells and fibroblasts was observed, however, following admixture of albumin with the vitreous. These findings suggest that vitreous contributes modulators that stimulate some functions of RPE cells that are believed to play a role in the pathogenesis of PVR. Invest Ophthalmol Vis Sci 30:1951-1957,1989

who possibly have less extensive damage of the blood-ocular barrier. It is not clear, however, whether the intravitreal mitogenic and chemotactic activity of patients with PVR,7 or of animals with experimental PVR,9 can be attributed solely to the presence of serum-derived factors or whether endogenous vitreous factors also contribute to an environment favoring RPE chemotaxis and proliferation. Mitogenic activity of bovine and porcine vitreous on RPE cells and fibroblasts has been reported,10 although others found no such effect." To define whether human vitreous contains modulators of RPE functions, we tested the effect of human vitreous on migration and proliferation of cultured human RPE cells under various experimental conditions.

Proliferative vitreoretinopathy (PVR) is characterized by the formation of contractile cellular membranes on the surface of the retina and in the vitreous. The cellular components of these membranes include fibroblasts from a wound site or from vascular areas of the optic nerve head, glial cells, retinal pigment epithelial (RPE) cells or combinations of these cell types.1"4 As these cells are not present normally in the vitreous, their migration and proliferation appear to play a role in the development of epiretinal membranes; however, the signals that initiate, maintain and eventually terminate epiretinal membrane formation in the eye are not understood. Since serum supplementation is necessary to grow RPE cells in culture, and since serum factors stimulate migration of RPE cells,5'6 it has been suggested that passage of serum factors into the vitreous could result in RPE chemotaxis and proliferation.7 This hypothesis is supported by the clinical observation that patients who have had multiple retinal reattachment procedures, and presumably greater damage to the bloodocular barrier, are at higher risk to develop PVR than are patients with uncomplicated retinal detachment,8

Materials and Methods Vitreous Collection For the isolation of vitreous, human eyes were obtained through the Lions Doheny Eye Bank. The average age of the donors was 69 years. The elapsed time between death of the donor and collection of the vitreous varied from 20 m in to 11 hr. The eyes were frozen in liquid nitrogen and bisected through the sclera approximately 1 mm posterior to the limbus; the anterior segment was then removed and the sclera, choroid and retina were peeled off the vitreous. Using this procedure, no microscopic evidence of contamination of the vitreous by other tissues was seen. The vitreous was then thawed, homogenized, centrifuged at 10,000 g for 30 min at 4°C, sterilized by filtration through a 0.2 /Ltm detergent-free, low protein binding filter (Gelman Sciences, Ann Arbor,

From the Department of Ophthalmology, University of Southern California School of Medicine, and Estelle Doheny Eye Institute, Los Angeles, California. * Present address: Universitats-Augenklinik Koln, Koln, West Germany. Supported by the Volkswagen Foundation, Hannover, West Germany, and in part by Grant EY-02061 from the National Eye Institute, National Institutes of Health, Bethesda, Maryland. Submitted for publication: August 5, 1987; accepted January 18, 1989. Reprint requests: Nino Sorgente, PhD, Doheny Eye Institute, 1355 San Pablo Street, Los Angeles, CA 90033.

1951

1952

INVESTIGATIVE OPHTHALMOLOGY 6 VISUAL SCIENCE / September 1989

MI) and stored at -90°C until used. Protein, as determined by the Bradford method,12 ranged from 440 to 1390 iig/m\ in the different lots of vitreous.

Vol. 30

142719, from Boerhinger Mannheim, Indianapolis, IN). Chemotaxis Assay

Retinal Pigment Epithelial Cell Isolation and Culture

For the isolation of RPE cells, eyes were cut circumferentially through the sclera approximately 1 mm posterior to the ora serrata; the vitreous was aspirated and the retina gently separated from the RPE cell layer, after which the eye cup was washed with Eagle's minimum essential medium (MEM, Irvine Scientific, Santa Ana, CA) and incubated at 37°C with 0.25% trypsin-0.02% EDTA (Irvine Scientific). After 1 hr the trypsin solution was aspirated from the eye cup and replaced with MEM supplemented with 20% fetal bovine serum (FBS) (HyClone, Logan, UT, or Gemini Bio-Products, Calabasas, CA). The RPE cells were released by gentle pipetting and transferred to a 75 cm2 tissue culture flask containing MEM supplemented with 20% FBS without antibiotics. The cells were cultured at 37°C in a humidified atmosphere of 5% CO2-95% air; the medium was changed every 4 days thereafter. When the cells had reached confluence, after 4-5 weeks, they were passaged by trypsinization. Third passage cells were used for all experiments. Human foreskin fibroblasts were obtained from the American Type Culture Collection (Rockville, MD) and cultured as described for RPE cells. Proliferation Assay

To determine the effect of human vitreous on cell growth, RPE cells were plated in 12-well cluster plates (Costar, Cambridge, MA) at an initial density of 1 X 104 cells/well in 2.0 ml of MEM supplemented with 5% FBS. To differentiate between the effect on cell proliferation and on cell attachment, the medium was changed 24 hr later and replaced with fresh medium containing the substances to be assayed. The cell number at 24 hr was used as the "0" time count; the number of cells at 24 hr was 70-80% of the number of cells seeded. The final cell number was determined after 14 days using a Coulter Counter (Coulter Electronics, Hialeah, FL). All experiments were done in triplicate and repeated at least three times. Human serum was obtained from Irvine Scientific (Type AB, Lot #500961210). Bovine serum albumin (BSA) was obtained from three different sources (Fraction V, Cat No. 81-003-3, Lot No. 412, from Miles Scientific, Naperville, IL; Fraction V, Cat No. A7638, Lot No. 116E 9390, from Sigma Chemical Co., St. Louis, MO; fatty acid-free, Cat No. 100062, Lot No.

RPE cell migration was examined using blind well chemotaxis chambers (Neuroprobe, Bethesda, MD). Confluent cells were dissociated with cold (4°C) 0.05% trypsin-0.02% EDTA (Irvine Scientific) for 3-5 min. The cells were centrifuged and washed one time in MEM supplemented by 10% FBS, resuspended in MEM and filtered through Nitex (Tatko, Monterey Park, CA) mesh screens 40 A*m in diameter to obtain a single cell suspension. The lower wells of the chemotaxis chambers contained MEM or MEMBSA (controls) with or without the various substances to be tested. An 8 /im pore-size polycarbonate chemotaxis filter (Nucleopore, Pleasanton, CA), which had been previously coated with gelatin (Sigma Chemical Co.), separated the upper and lower wells. Chemotaxisfilterswere coated with gelatin by boiling them in a 0.5 mg% solution of gelatin, according to the method of Postlethwaite, Snyderman and Kang.13 RPE cells (5 X 104) were placed in the upper wells and the chambers incubated at 37°C in a humidified atmosphere of 95% air-5% CO2. After 6 hr the filters were removed from the chambers and the nonmigrated cells removed by wiping the upper surface of the filters with a cotton-tipped swab. The filters were fixed in half-strength Karnovsky's fixative and stained with Richardson-blue. The number of cells migrating through the filters was counted in ten XI000 microscope fields. Triplicate samples were assayed in all experiments and each experiment was repeated at least three times. Statistical analysis of our in vitro experiments was based on the log of the cell numbers, since the log values have a more normal distribution than do the cell numbers. An analysis of variance test was done using the student-Newman-Keuls procedure for comparing individual means. Results Proliferation

In the absence of serum, human vitreous had no significant effect on RPE cell proliferation (Fig. 1, left), whether collected 20 to 60 min or 4 to 11 hr postmortem. FBS up to a concentration of 25% stimulated RPE cell proliferation with a dose-response relationship; at a concentration of 50%, however, FBS exerted a strongly inhibitory effect on RPE proliferation (Fig. 1, right). The addition of vitreous to

No. 9

Fig. 1. Left, Effect of human vitreous on the proliferation of human RPE cells. Bars are the means ± standard deviation of the means of three separate experiments. Right, The effect of human vitreous in the presence of FBS on the proliferation of human RPE cells. Bars indicate the means ± standard deviation of the means of a representative experiment. Cell numbers in wells containing vitreous-serum admixtures were significantly higher (P < 0.0001) than those from wells containing serum only. Vitreous 25% and BSS (balanced salt solution) 25% indicate the media contained 25% of these by volume.

EFFECTS OF VITREOUS ON RETINAL PIGMENT EPITHELIUM / Kirchhof er ol

^ B Vitreous

Vitreous 25% + Serum

rziBSS

BSS 25% + Serum

10%

0.5%

25% 50%

1% 2%

3%

5% 10% 25% 50%

Serum

cells growing in the presence of serum increased cell proliferation in a remarkable manner; even when serum was present at concentrations as low as 0.5%, vitreous at a concentration of 25% stimulated cell proliferation by more than 200%, suggesting a synergistic effect of serum and vitreous. As noted in the presence of 50% FBS, RPE cell proliferation was inhibited; however, the addition of vitreous again stimulated cell proliferation (Fig. 1, right). RPE cells culFig. 2. Left, Effect of human vitreous in the presence of BSA on the proliferation of human RPE cells. Bars are the means ± standard deviation of triplicate wells of a representative experiment. Cell numbers in wells containing vitreous were significantly higher than cell numbers in wells containing BSS (balanced salt solution) instead (P < 0.0001). Right, Effect of vitreous in the presence of BSA on the proliferation of human foreskin fibroblasts (HFF). Bars are mean values ± standard deviation of triplicate wells of a representative experiment. Cell numbers in wells containing vitreous were significantly higher than those without vitreous (P < 0.0001).

1953

tured in human serum exhibited the same qualitative response as did those cultured in FBS. Since vitreous in the presence of 50% serum, which by itself is inhibitory to cell proliferation, increased cell proliferation more than in any other serum concentration, it would appear that the serum provided factors other than growth factor(s). To investigate whether the stimulation of RPE cell proliferation by vitreousserum mixture requires the presence of serum growth

•

BSA 2g% + BSS

•

BSA 2g%+VITREOUS

§3 VITREOUS HFF

0.1

0.5

1

2

4

8

CONCENTRATION OF BSA IN g%

10% 25% 50% CONCENTRATION OF VITREOUS OR BSS IN VOL%

1954

Vol. 30

INVESTIGATIVE OPHTHALMOLOGY 6 VISUAL SCIENCE / Seprember 1989

serum or BSA stimulates RPE proliferation over a period of at least nine days (Fig. 3). The effect of vitreous is species-specific, since bovine vitreous in the presence of 5% FBS stimulated the proliferation of bovine but not human RPE cells (Fig. 4). The active chemotactic and mitogenic agents in vitreous are unknown; as can be seen in Figure 5 (lanes 3-5), vitreous is a very heterogeneous tissue; similarly, the albumin preparations used in this study showed many impurities (Fig. 5, lanes 6-8). Chemotaxis

DAYS

Fig. 3. Effect of vitreous in the presence of fetal bovine serum (FBS) or bovine serum albumin (BSA) on the proliferation of human retinal pigment epithelial cells in culture. Analysis of variance showed that cell numbers in the groups treated with bovine serum albumin plus vitreous and those treated with fetal bovine serum plus vitreous were significantly different from the groups treated only with BSA, FBS, or vitreous. Bars are the means ± standard deviation of the means of three separate experiments.

Human vitreous increased the migration of human RPE cells even in the absence of FBS. Maximal stimulation was observed in the presence of 100% vitreous; the addition of serum or 2 g% BSA to diluted vitreous increased cell migration to the level of undiluted vitreous (Fig. 6, left); surprisingly, however, 2 g% BSA also increased migration by the same extent as did vitreous and BSA. Serum alone also increased RPE cell migration, in agreement with the report of Campochiaro et al.5 To determine whether vitreous stimulated chemotaxis, chemokinesis or both, a checkerboard assay was performed. The results of this experiment (Fig. 6, right) showed that vitreous in-

Bovine Vitreous CD

• Human RPE

O

• Bovine RPE factor(s), serum was replaced by various concentrations of serum albumin equivalent to 1.5% to 95% of the concentrations provided by 50% serum. Figure 2 (left) shows that vitreous in the presence of albumin was nearly as effective a mitogen as in the presence of serum, and that increasing the concentration of albumin up to 2 g% increased the effect of vitreous; albumin alone had no effect. Albumin from two other sources was equally effective (data not shown). Cell proliferation was stimulated slightly less with albumin-vitreous than with serum-vitreous, suggesting that the effect of the serum-vitreous represents the sum of serum-derived and vitreous-derived mitogen ic activity. Figure 2 (right) shows that human vitreous also stimulates the proliferation of human fibroblasts, but only in the presence of BSA and in a dose dependent manner; as for RPE cells, the maximum effect was reached at a 2 g% concentration of albumin. Addition of vitreous to media at time "0" in the presence of

CD _Q

§1.5

* I

0)

0.5

0

FBS 10% 25% 50% Concentration of Vitreous in vol% Fig. 4. Effect of bovine vitreous in the presence of 5% FBS on the proliferation of human and bovine RPE cells. Bars are means ± standard deviation of triplicate wells of a representative experiment. Bars marked with an asterisk indicate a significant difference in cell number compared with the FBS groups (P < 0.0001).

1955


No. 9

Discussion

KD 200.0116.397.4~ 66.242.731.021.514.4 1

6

7

8

Fig. 5. Sodium dodecyl sulfate polyacrylamide-gel electrophoresis of human vitreous and BSA. Lane 1, standards high MW; lane 2, standards low MW; lane 3, vitreous from a 69-year-old male obtained 20 hr postmortem; lane 4, vitreous from a 51-year-old female obtained 16 hr postmortem; lane 5, vitreous from a 48-yearold male obtained 18 hr postmortem; lane 6, BSA (Boehringer Mannheim), fatty acid-free; lane 7, BSA (Miles Laboratories, Pentex®); lane 8, BSA (Sigma Chemical Co.).

creases both chemotaxis and chemokinesis. A 5-fold increase in migrating cells was observed when the upper and lower chambers contained 50% vitreous, whereas an 8.6-fold increase in migrating cells occurred when 50% vitreous was present in the lower chamber only. Fig. 6. Left, Effect of human vitreous, human serum and BSA on the migration of human RPE cells. Bars are the means of migrated cells ± standard deviation from three membranes from each of three separate experiments. Analysis of variance showed that all experimental groups were significantly different from the MEM control (P < 0.05). Right, Migration of human RPE cells in response to varying concentrations of human vitreous in MEM. The numbers are means of migrated RPE cells ± standard deviation in ten XI000 fields from three separate experiments. Diagonal lines demarcate cell migration in the absence of a concentration gradient.

To our knowledge, there are no reports in the literature of the effect of postmortem human vitreous on the proliferation and migration of human RPE cells, even though in vitro studies of the effects of bovine vitreous or vitreous extracts on human and animal RPE cells have been reported. Lutty et al" demonstrated that various bovine vitreous fractions had no effect on the proliferation of human RPE cells; however, whole vitreous was not tested. Wiedemann et al10 showed that unfractionated bovine vitreous stimulated the proliferation of porcine RPE as well as the proliferation of bovine and lapine fibroblasts. Since the different results obtained by these two investigators could be due to species differences, we tested bovine and human vitreous on bovine and human RPE cell proliferation. Our results show that bovine vitreous in the presence of 5% FBS stimulated the proliferation of bovine RPE cells, but not of human RPE cells; human vitreous, on the other hand, in the presence of FBS stimulated the proliferation of human RPE cells but not bovine RPE cells (data not shown). Thesefindingsindicate that the species specificity lies not only with the effector, vitreous, but also within the target RPE cells. These results emphasize the importance of species variability when comparing results from different laboratories. The results presented here suggest two conclusions. The first is that vitreous contains a growth factor(s) specific for RPE cells, and that the RPE exhibits specificity of response

• MEM • El ^ ^ ^

MEM + vitreous FBS 1 % FBS 1%+vitreous BSA 2g%-hvitreous BSA 2g% Vitreous concentration 0%

140 120

0%

x 100 O o 80 o

12N 23 ~2 \

4/-6

en

40

O

50%

20 0

50%

28

\ o c o o

60

=*=

25%

0 20 50100 0 20 20 0

Concentration of vitreous in vol%

39N 28 \f/- S

103 73\

\

- 5

1956

INVESTIGATIVE OPHTHALMOLOGY & VISUAL SCIENCE / Seprember 1989

to vitreous. The second conclusion to be drawn from these studies is that human RPE cells, as well as human fibroblasts, do not respond to human vitreous unless albumin (or serum) is present, but that the majority of the in vitro proliferative and chemotactic activity of vitreous-serum mixtures is a property of vitreous. The finding that human RPE cells and fibroblasts in culture proliferate under serum-free conditions when exposed to BSA is unexpected, but is confirmed by multiple experiments that yielded a high statistical significance (P < 0.0001). The observation may be important because it reveals a level of regulation of cell proliferation previously unknown. The mechanism of action of albumin is unknown and needs further exploration. We speculate that the target cell does not recognize the vitreous growth factors), and that albumin mediates the uptake or binding of the growth factor to the cell receptors, as has been shown in other instances.14"17 Whether this action is a property specific to albumin or whether other proteins can be substituted for albumin remains to be investigated. That this effect is due to an impurity in the albumin preparations is certainly a consideration in view of the many impurities present (Fig. 5); however, since albumin alone had no effect on RPE cell proliferation, we assume that the mitogenic effect is a property of vitreous. It is also possible that the increased proliferation may be the result of synergism between a factor in the vitreous and an additional factor in the albumin preparation. Human vitreous contains not only mitogens for human RPE cells but also factors that mediate the migration of RPE cells. Campochiaro et al18 reported that rabbit vitreous only slightly stimulated migration of human RPE cells. The discrepancy between our results and those of Campochiaro and colleagues may again reflect a difference in specificity of response, as in the case of RPE cell proliferation. Campochiaro et al7 also studied the effect of vitreous from patients with PVR on human RPE cell migration, and attributed the increased migration that they observed to serum components in the vitreous; they did not, however, use normal vitreous as a control. Our results show that normal human postmortem vitreous is more chemotactic than is bovine or human serum, since the stimulation of migration caused by 1% FBS is equal to that induced by 20% human vitreous. Since serum contains 60-80 mg protein/ml, and vitreous 0.4 to 1 mg protein/ml, it is evident that, on a protein basis, vitreous is a more effective chemotactic agent than is FBS. The stimulation of RPE migration by albumin was surprising. We do not know the mechanism, but it may be similar to that which accounts for polymor-

Vol. 30

phonuclear leukocyte migration, that is, migration may be stimulated by reduction of the adhesion of the cells to the substratum.19 It is also possible that the migration promoting activity of albumin may be due to some impurity in the BSA preparations. Our in vitro results are consonant with clinical observations on the development of PVR. The appearance of pigmented RPE cells in the vitreous ("tobacco dust") may occur consequent to rhegmatogenous retinal detachment because they become exposed to the vitreous. However, these cells will not proliferate in the vitreous unless there is a break in the blood-ocular barrier that allows serum, including albumin, access to the vitreous. Depending on the amount of serum in the vitreous, and depending on the duration of blood-ocular barrier breakdown, cell proliferation may or may not proceed to reach a critical number of cells20 needed to form epiretinal membranes that subsequently exert enough tractional force to detach the retina. This hypothesis would explain the higher incidence of PVR in eyes with multiple surgical interventions, which favor blood-ocular barrier breakdown,21 compared with eyes having uncomplicated retinal detachments.8 Key words: human RPE cells, human vitreous, albumin, serum, in vitro, migration, proliferation Acknowledgments The authors thank William Barlow, PhD, Biometry Division, Department of Preventive Medicine, University of Southern California, for statistical evaluation of the results.

References 1. Machemer R and Laqua H: Pigment epithelium proliferation in retinal detachment (massive periretinal proliferation). Am J Ophthalmol 80:1, 1975. 2. Machemer R, van Horn D, and Aaberg TM: Pigment epithelial proliferation in human retinal detachment with massive periretinal proliferation. Am J Ophthalmol 85:181, 1978. 3. Newsome DA, Rodrigues MM, and Machemer R: Human massive periretinal proliferation: In vitro characteristics of cellular components. Arch Ophthalmol 99:873, 1981. 4. Radtke ND, Tano Y, Chandler D, and Machemer R: Simulation of massive periretinal proliferation by autotransplantation of retinal pigment epithelial cells in rabbits. Am J Ophthalmol 91:76, 1981. 5. Campochiaro PA, Jerdan JA, and Glaser BM: Serum contains chemoattractants for human retinal pigment epithelial cells. Arch Ophthalmol 102:1830, 1984. 6. Campochiaro PA and Glaser BM: Platelet-derived growth factor is chemotactic for human retinal pigment epithelial cells. Arch Ophthalmol 103:576, 1985. 7. Campochiaro PA, Jerdan JA, Glaser BM, Cardin A, and Michels RG: Vitreous aspirates from patients with proliferative vitreoretinopathy stimulate retinal pigment epithelial cell migration. Arch Ophthalmol 103:1403, 1985.

No. 9


8. Rachal WF and Burton TC: Changing concepts of failures after retinal detachment surgery. Arch Ophthalmol 97:480, 1979. 9. Grierson I, Boulton M, Hiscott P, Hitchins C, Gilbert D, and McLeod D: Human retinal pigment epithelial cells in the vitreous of the owl monkey. Exp Eye Res 43:491, 1986. 10. Wiedemann P, Ryan SJ, Novak P, and Sorgente N: Vitreous stimulates proliferation offibroblastsand retinal pigment epithelial cells. Exp Eye Res 41:619, 1985. 11. Lutty GA, Mello RJ, Chandler C, Fait C, Bennett A, and Patz A: Regulation of cell growth by vitreous humour. J Cell Sci 76:53, 1985. 12. Bradford MM: A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal Biochem 72:248, 1976. 13. Postlethwaite AE, Snyderman R, and Kang AH: The chemotactic attraction of humanfibroblaststo a lymphocyte-derived factor. J Exp Med 144:1188, 1976. 14. Blitzer BL and Lyons L: Enhancement of Na+-dependent bile acid uptake by albumin: Direct demonstration in rat basolateral liver plasma membrane vesicles. Am J Physiol 249:G34, 1985.

1957

15. Peters T: Serum albumin. In The Plasma Proteins, Putnam FW, editor. New York, Academic Press, 1975, pp. 133-181. 16. Weisiger R, Gollan J, and Ockner R: Receptor for albumin on the liver cell surface may mediate uptake of fatty acids and other albumin-bound substances. Science 211:1048, 1981. 17. Weisiger R, Gollan J, and Ockner R: The role of albumin in hepatic uptake processes. In Progress in Liver Diseases, Vol. VII, Popper H and Schaffner F, editors. New York, Grune & Stratton, 1982, pp. 71-85. 18. Campochiaro PA, Bryan JA III, Conway BP, and Jaccoma EH: Intravitreal chemotactic and mitogenic activity: Implication of blood-retinal barrier breakdown. Arch Ophthalmol 104:1685, 1986. 19. Valerius NH: Chemotaxis, spreading and oxidative metabolism of neutrophils: Influence of albumin in vitro. Acta Pathol Microbiol Immunol Scand 91:43, 1983. 20. Fastenberg DM, Diddie KR, Dorey K, and Ryan SJ: The role of cellular proliferation in an experimental model of massive periretinal proliferation. Am J Ophthalmol 93:565, 1982. 21. Sears ML: Aphakic cystoid macular edema: The pharmacology of ocular trauma. Surv Ophthalmol 28:525, 1984.