- Email: [email protected]
Progesterone induces vascular endothelial growth factor on retinal pigment epithelial cells in culture
PI1 SOO24-3205(96)00253-6
ELSEVIER
Life Sciences, Vol. 59, No. 1, pp. 21-25,1996 copyright 0 1996 Eavier scie.nce Inc. Printed in the USA. All rights resewed lmm-32Qs/% $l.5.00 t .al
PROGESTERONE INDUCES VASCULAR ENDOTIIELIAL ON RETINAL PIGMENT EPITHELIAL CELLS
Hirohito
Sonel,
Mitsuya Hanatan?,
GROWTH IN CULTURE
FACTOR
, Shinichi Kondo2, Okuda l* , Yasushi Kawakamil Katsuhiko Matsu&, Hideo Suzuki2 and Kamejiro Yamashital
Yukichi
1Division of Endocrinology and Metabolism, Department of Internal Medicine, University of Tsukuba, Tsukuba, 305 Japan 2Bioscience Research Department, Tsukuba Research laboratory, Toagosei Co. Ltd. Tsukuba, 305 Japan (Received in final form April 29,19%)
Summary
Diabetic retinopathy is known to frequently deteriorate during pregnancy but the cause remains obscure. Vascular endothelial growth factor (VEGF), also known as vascular permeability factor (VPF), is a potent vascular endothelial cell mitogen which is mainly up-regulated by hypoxia, and is closely associated with the development and progression of diabetic retinopathy. To examine the influence of the drastic hormonal alterations during pregnancy on the worsening of diabetic retinopathy, we examined the effects of estradiol (Ez) and progesterone (Pa) on the production of “‘?F:/VPF in bovine retinal pigment epithelial cells in culture. The VEGF/VPF production was significantly elevated (214.5+28.3 nglg protein, PtO.01) by 48 h of esposure to a high concentration of P_t(10 PM), which is still within the physiological range during pregnancy, compared to that of the control group (147.7+17.9 rig/g protein). However, E: significantly stimulated the production of VEGF/VPF only at concentrations (100 PM) much higher than normally encountered during pregnancy. These two hormones were not observed to have a synergistic effect, at least at physiological concentrations. As the increase in serum Pa levels during pregnancy is reported to be greater in pregnant diabetic patients with progressive retinopathy, our findings suggest that P4 may contribute to the worsening of diabetic retinopathy during pregnancy by up-regulating intraocular VEGF levels. Key Worde diabetes melllitus, diabetic retinopathy, vascular endothelial growth factor, vascular permeability factor, estrogen, progesterone, retinal pigment epithelial cell While diabetic retinopathy frequently deteriorates during pregnancy, the exact reason remains obscure (l-3). Klein et al. (1) found an odds ratio of 2.3 for pregnancy as a risk factor for the worsening of diabetic retinopathy. One of the most extreme endocrinological alterations throughout pregnancy is a marked elevation of both estrogen and progesterone (3). Furthermore, it is reported that the serum levels of progesterone during pregnancy are significantly higher in diabetic patients compared with normal pregnant subjects (4, 5). Moreover, pregnant diabetic patients with progressive retinopathy showed progesterone and estradiol at or above the upper limit of the normal range for pregnancy (5). However, neither estradiol nor progesterone had an effect on cell multiplication (6).
* To whom correspondence
should be addressed
22
VEGF Induction on Retina by Progesterone
Vol. 59, No. 1, 1996
Vascular endothelial growth factor (VEGF), also known as vascular permeability factor (VPF), is an endothelial-specific potent stimulator of neovascularization which also markedly promotes vascular permeability (7). It is related to platelet-derived growth factor (PDGF) in its structure (7). Recent studies (8, 9) provide strong evidence that VEGF plays important roles in the development and progression of diabetic retinopathy as do insulin-like growth factor-I, basic fibroblast growth factor (bFGF) and PDGF (10, 11). Although VEGF is known to be up-regulated primarily by hypoxia (12), some other factors (e.g. epidermal growth factor (13), prostaglandin EI and E2 (14), transforming growth factor-p (15, 16), interleukin-lp (17)) are also reported to induce VEGF production. Recently, the expression of VEGF mRNA was induced by estradiol in cell lines of rat uterine carcinoma, and its induction by estradiol, estriol or progesterone was also observed in the rat uterus in vivo (16, 18). However, it is not clear whether VEGF is up-regulated by estrogen or progesterone in cells derived from non-tumor, non-reproductive organs. To clarify whether VEGF is induced by a high concentration of estrogen or progesterone on retinal cells, and whether there are any synergistic effects between these hormones on VEGF secretion, we examined the VEGF production of cultured retinal cells under various concentrations of estrogen and progesterone.
Materials
and
methods
Cell Culture Primary cultures of bovine retinal pigment epithelial cells (RPE) were established as previously described with slight modifications (19). Briefly, the globes were bisected circumferentially 3 mm posterior to the limbus and the vitreous removed. The remaining eyecups were incubated with 0.25% trypsin in 0.02% EDTA (Gibco Laboratories, Grand Island, NY) for 15 min at 37”~. Removed RPE was transferred to collagen-coated flasks (Coming, New York, NY) and grown in Dulbecco’s modified Eagle’s medium (Gibco) supplemented with 10% fetal calf serum (Gibco) and 1% guinea pig serum (IBL, Gunma, Japan) under 5% CO2 at 37”~ and relative humidity of 95%. The media were changed every 3 days. At confluence, cells were subcultured using 0.25% trypsin in 0.02% EDTA (Gibco). Cells at the third passage were used for the experiment. Cells were identified histochemically by anti-cytokeratin antibody (Sigma Chemical Co., St. Louis, MO). VEGF production on high estrogen and/or progesterone concentrations RPE was re-plated in eleven plastic plates with six non-coated wells (Becton Dickinson Labware, Lincoln Park, NJ) and grown until confluency. The previously mentioned RPE medium but containing different concentrations of 17-OH estradiol p (Ez) (Sigma) (0.01, 1, 10, 100 FM), progesterone (Pd) (Sigma) (1, 2, 10,100 PM) and their combinations (E2 0.01 pM+Pd 1 PM, E2 1 ~M+PJ 10 PM), was prepared for each plate, while original RPE medium was supplied to a plate for the control group. As these agents are water-soluble products containing dextrin, dextrin was added to each well to make the amount of dextrin contained in each well equal. After incubation for 48 h, the conditioned media (1.5 ml per well) were collected and centrifuged. The supernatant was stored at -70 “C until assay. Determination of VEGF levels in cultured media The VEGF concentrations in the conditioned media were measured by our newly improved highly sensitive enzyme-linked immunosorbent assay (ELISA) using rabbit anti-human VEGF/VPF polyclonal antibody for both capture (solid phase) and secondary (enzyme-labeled) antibodies (20). This immunoassay detects all known isoforms of VEGF. The limit of detection for the assay, defined as +2SD above the zero standard, is 1 pg/ml. The intra-assay coefficient of variation was less than 6.1%. All the conditioned media were assayed after concentrating them to one-fourth of the original volume by vacuum centrifugation. Neither of these concentrating processes nor the concentrations of E 2, P4 or dextrin interfered with our VEGF assay (data not shown). All the measurements were corrected by protein content in each well measured by the method of Lowry et al. (21). All the results are expressed as mean&D and analysis of variance (ANOVA) was used for statistical analysis, and PcO.05 was taken as the level of significance.
VEGF Induction on Retina by Progesterone
Vol. 59, No. 1,1996
23
Results
In the Ez groups, only the VEGF concentrations at 100 PM of & (204.6k20.3 nglg protein, PcO.05) showed significant elevation, while the 0.01 PM (152.8+18.4 rig/g protein), 1 PM (171.7k22.2 rig/g protein) and 10 PM (153.5t23.7 rig/g protein) groups did not show significant changes compared to the control group (147.7k17.9 nglg protein) (Fig. 1). In the Pa group, 10 PM (214.51283 rig/g protein, PcO.01) and 100 PM (211.2&27.1 rig/g protein, P
E2b-W P4(lW
0 0 (control)
0.01 0
1 0
10 0
0 1
loo 0
:
0 10
0 loo
0.01 1
1 10
Fig. 1
Vascular endothelial growth factor levels in conditioned medium of bovine retinal pigment epithelial cells stimulated with various concentrations of estradiol and/or progesterone (N=6, Mean&D, * P&.0.5, ** PcO.01 compared with control group, NS; not significant)
Discussion
The reason for the worsening of diabetic retinopathy during pregnancy, especially pathological neovascularization or exudative changes frequently observed in patients with longterm diabetes, remains obscure. Too rapid normalization of blood glucose levels (22, 23) or hypertension associated with pregnancy (24) may be involved but these factors do not explain the entire phenomenon (12). Our findings demonstrated that the induction of VEGF by high concentrations of estradiol or progesterone is not limited to tumor cells and reproductive organs. Although the induction of VEGF by E2 at 0.01 PM was observed in HECl-A cells, a human endometrial adenocarcinoma cell line, it was seen only at concentrations far above the physiological range (100 PM) in RPE. This might be due to variability in the number of estrogen receptors that each cell possesses. (The distribution of estrogen receptors is not limited to the target tissues (25)) On the other hand, our findings indicated significant up-regulation of VEGF by physiologically elevated concentrations of P4 during pregnancy. During pregnancy, upregulation of P4, but not b, has been shown to contribute to the worsening of diabetic retinopathy through up-regulated VEGF. This hypothesis is supported by the clinical observations of Larinkari et al (5) who reported that, during the second trimester, significantly higher levels of serum progesterone but not estradiol were found in pregnant diabetic patients with retinopathy compared to those without retinopathy. The reasons why pathological retinal proliferation occurs
VEGF Inductionon Retina by Progesterone
24
Vol. 59, No. 1, 1996
only in diabetic but not in normal pregnant subjects may be the requirement of pre-existing background retinopathy and/or the need for hypoxia to induce sufficient up-regulation of VEGF. Although most reports determining stimulating factors of VEGF induction are based on elevation of its mRNA levels (12- 15), we confirmed such induction using the protein levels. Since our VEGF protein assay uses anti-human VEGF antibody, our measurements represent relative values, not the absolute values of the VEGF concentration in the conditioned medium. Four isoforms of VEGF are produced from a single gene by alternative splicing (7). The four isoforms each have 121, 165, 189,206 amino acids after removal of the signal peptide, and the former two are released from the producing cells, while the latter two bind to the extracellular matrix (26). Our assay using the antibody against the 121 amino acid form of VEGF detects all of those isoforms. Retinal pigment epithelium forms a single layer of cells which compose the outermost layer of the neural retina. Its contribution to diabetic retinopathy has been underestimated because of its location. However, the retinal pigment epithelium is known to form an integral part of the outer blood-retinal barrier, and its breakdown is known as the earliest pathological change in diabetic retinopathy (27). In a diabetic rat model, this breakdown was found to be due to alteration of membrane permeability rather than a loss of tight junctions (28). RPE-derived VEGF may play some roles in this process through autocrine pathways, because VEGF/VPF is known to mai kedly promote vascular permeability (7), as its name indicates. Moreover, RPE produces mitogenic factors for retinal microvascular cells (29) including bFGF (30), that has a synergistic effect with VEGF (3 1), and evidence that RPE contributes to the proliferative diabetic retinopathy (PDR) membrane has been obtained by ultrastructural investigation by Hamilton et al. (32). Furthermore, Hiscott et al. (33) histochemically showed in PDR patients that RPE is contained (520%) in combined traction rhegmatogenous retinal detachment membrane. These observations strongly suggest that RPE migrates through retinal breaks to access the PDR membrane and could contribute to its progression by secreting some angiogenic factors. In fact, the PDGF A chain was confirmed to be expressed in RPE of the PDR membranes (11). Half-maximal stimulation of endothelial cell growth is obtained at 100-150 pg/ml of VEGF (7), which is approximately the same concentration range of the RPE-conditioned media obtained in our study.
Acknowledement The authors thank Ms.Tomoe Matsumoto of Toagosei Co. Ltd. for her excellent technical assistance, and Chikusei Meat Center, All Japan Stockbreeding Public Corporation for their kind supply of bovine eyes. Thanks are also extended to Dr. Yasuo Sekine and Prof. Sachiko Honmura of the Department of Ophthalmology, University of Tsukuba for their encouragement.
References 1. 2.
B. KLEIN, S. MOSS and R. KLEIN, Diabetes Care. 12 34-40 (1990). K. ELMAN, R. WELCH, R. FRANK, G. GOYERT and R. SOKOL, Obstet Gynecol. 75
3.
L. JOVANOVIC-PETERSON and C. PETERSON, Clin Obstet Gynecol. 34 516-525 (1991). M. BOOTH and T. EL GARF, J Obstet Gynaecol Br Comwth. a 768-776 (1974). J. LARINKARI, L. LAATIKAINEN, T. RANTA, P. MORONEN, K. PESONEN and T. LAATIKAINEN, Diabetologia. 22 327-332 (1982). E. CORVAZIER, E. DUPUY, A. DOSNE and J. MACLOUF, Thromb Res. 34 303-310 (1984). N. FERRARA, K. HOUCK, L. JAKEMAN and D. LEUNG, Endocr Rev. 13 18-32 (1992). L. AIELLO, R. AVERY, P. ARIGG, B. KEYT, H. JAMPEL, S. SHAH, L. PASQUALE, H. THIEME, M. IWAMOTO, J. PARK, H. NGUYEN, L. AIELLO, N. FERRARA and G. KING, N Engl J Med. 331 1480-1487 (1994).
119-127 (1990).
4. 5. 6. 7. 8.
Vol. 59, No. 1, 1996
9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 2.5. 26. 27. 28. 29. 30. 31. 32. 33.
VEGF
Induction on Retina by Progesterone
A.P. ADAMIS, J.W. MILLER, M.T. BERNAL, D.J. D’AMICO, J. FOLKMAN, T.K. YE0 and K.T. YEO, Am J Ophthalmol. 118 445450 (1994). M. BARINAGA, Science. 267 452-453 (1995). S. ROBBINS, R. MIXON, D. WILSON, E. HART, J. ROBERTSON, I. WESTRA, S. PLANCK and J. ROSENBAUM, Invest Ophthalmol Vis Sci. 35 3649-3663 (1994). D. SHWEIKI, A. ITIN, D. SOFFER and E. KESHET, Nature. 359 843-845 (1992). C. GOLDMAN, J. KIM, W. WONG, V. KING, T. BROCK and G. GILLESPIE, Mol Biol Cell. 4 121-133 (1993). S. HARADA, J. NAGY, K. SULLIVAN, K. THOMAS, N. ENDO, G. RODAN and S. RODAN, J Clin Invest. 93 2490-24% (1994). L. PERTOVAARA, A. KAIPAINEN, T. MUSTONEN, A. ORPANA, N. FERRARA, 0. SAKSELA and K. ALITALO, J Biol Chem. 269 6271-6274 (1994). B. CULLINAN and R. KOOS, Endocrinology. 133 829-837 (1993). J. LI, M. PERRELLA, J. TSAI, S. YET, C. HSIEH, M. YOSHIZUMI, C. PATTERSON, W. ENDEGE, F. ZHOU and M. LEE, J Biol Chem. 270 308-3 12 (1995). J. CHARNOCK, A. SHARKEY, W. RAJPUT, D. BURCH, J. SCHOFIELD, S. FOUNTAIN, C. BOOCOCK and S. SMITH, Biol Reprod. 48 1120-l 128 (1993). A. ADAMIS, D. SHIMA, K. YEO, T. YEO, L. BROWN, B. BERSE, P. D’AMORE and J. FOLKMAN, Biochem Biophys Res Commun. 193 63 1-638 (1993). M. HANATANI, Y. TANAKA, S. KONDO, I. OMORI and H. SUZUKI, Biosci Biotech Biochem. 59 1958-1959 (1995). 0. LOWRY, N. ROSENBROUGH, A. FARR and J. RANDALL, J Biol Chem. &I 265 275 (1951). K. DAHL-JORGENSEN, 0. BRINCHMANN-HANSEN, K. HANSSEN, L. SANDVIK, F. AAGENSAES and A.D. GROU, Br Med J. 290 811-815 (1985). L. LAATIKAINEN, K. TERAMO, H.H. HIETA, V. KOIVISTO and R. PELKONEN, Acta Med Stand. 221 367-376 (1987). B. ROSENN, M. MIODOVNIK, G. KRANIAS, J. KHOURY, C. COMBS, F. MIMOUNI, T. SIDDIQI and M. LIPMAM, Am J Obstet Gynecol. 166 1214-1218 (1992). D. CIOCCA and L. ROIG, Endcr Rev. 16 35-62 (1995). K. HOUCK, D. LEUNG, A. ROWLAND, J. WINER and N. FERRARA, J Biol Chem. 267 26031-26037 (1992). S. WALTAM, C. OESTRICH, T. KRUPIN, S. HANISH, S. PATZAN, J. SANTIAGO and C. KILO, Diabetes. 27 85-87 (1978). R. CALDWELL, S. SLAPNICK and B. MCLAUGHLIN, Curr Eye Res. 4 215-227 (1985). H. WONG, M. BOULTON, P. CLARK, M. BAYLY and J. MARSHALL, Eye. 1. 754756 (1987). L. BOST, A. AOTAKI-KEEN and L. HJELMELAND, Exp Eye Res. 55 727-734 (1992). F. GOTO, K. GOTO, K. WEINDEL and J. FOLKMAN, Lab Invest. @ 508-517 (1993). C. HAMILTON, D. CHANDLER, G. KLINTWORTH and R. MACHMER, Am J Ophthalmol. 94 479-488 (1982). P. HISCOTT, R. GRAY, I. GRIERSON and Z. GREGOR, Br J Ophthalmol. 2s 219-222 (1994).
25
ELSEVIER
Life Sciences, Vol. 59, No. 1, pp. 21-25,1996 copyright 0 1996 Eavier scie.nce Inc. Printed in the USA. All rights resewed lmm-32Qs/% $l.5.00 t .al
PROGESTERONE INDUCES VASCULAR ENDOTIIELIAL ON RETINAL PIGMENT EPITHELIAL CELLS
Hirohito
Sonel,
Mitsuya Hanatan?,
GROWTH IN CULTURE
FACTOR
, Shinichi Kondo2, Okuda l* , Yasushi Kawakamil Katsuhiko Matsu&, Hideo Suzuki2 and Kamejiro Yamashital
Yukichi
1Division of Endocrinology and Metabolism, Department of Internal Medicine, University of Tsukuba, Tsukuba, 305 Japan 2Bioscience Research Department, Tsukuba Research laboratory, Toagosei Co. Ltd. Tsukuba, 305 Japan (Received in final form April 29,19%)
Summary
Diabetic retinopathy is known to frequently deteriorate during pregnancy but the cause remains obscure. Vascular endothelial growth factor (VEGF), also known as vascular permeability factor (VPF), is a potent vascular endothelial cell mitogen which is mainly up-regulated by hypoxia, and is closely associated with the development and progression of diabetic retinopathy. To examine the influence of the drastic hormonal alterations during pregnancy on the worsening of diabetic retinopathy, we examined the effects of estradiol (Ez) and progesterone (Pa) on the production of “‘?F:/VPF in bovine retinal pigment epithelial cells in culture. The VEGF/VPF production was significantly elevated (214.5+28.3 nglg protein, PtO.01) by 48 h of esposure to a high concentration of P_t(10 PM), which is still within the physiological range during pregnancy, compared to that of the control group (147.7+17.9 rig/g protein). However, E: significantly stimulated the production of VEGF/VPF only at concentrations (100 PM) much higher than normally encountered during pregnancy. These two hormones were not observed to have a synergistic effect, at least at physiological concentrations. As the increase in serum Pa levels during pregnancy is reported to be greater in pregnant diabetic patients with progressive retinopathy, our findings suggest that P4 may contribute to the worsening of diabetic retinopathy during pregnancy by up-regulating intraocular VEGF levels. Key Worde diabetes melllitus, diabetic retinopathy, vascular endothelial growth factor, vascular permeability factor, estrogen, progesterone, retinal pigment epithelial cell While diabetic retinopathy frequently deteriorates during pregnancy, the exact reason remains obscure (l-3). Klein et al. (1) found an odds ratio of 2.3 for pregnancy as a risk factor for the worsening of diabetic retinopathy. One of the most extreme endocrinological alterations throughout pregnancy is a marked elevation of both estrogen and progesterone (3). Furthermore, it is reported that the serum levels of progesterone during pregnancy are significantly higher in diabetic patients compared with normal pregnant subjects (4, 5). Moreover, pregnant diabetic patients with progressive retinopathy showed progesterone and estradiol at or above the upper limit of the normal range for pregnancy (5). However, neither estradiol nor progesterone had an effect on cell multiplication (6).
* To whom correspondence
should be addressed
22
VEGF Induction on Retina by Progesterone
Vol. 59, No. 1, 1996
Vascular endothelial growth factor (VEGF), also known as vascular permeability factor (VPF), is an endothelial-specific potent stimulator of neovascularization which also markedly promotes vascular permeability (7). It is related to platelet-derived growth factor (PDGF) in its structure (7). Recent studies (8, 9) provide strong evidence that VEGF plays important roles in the development and progression of diabetic retinopathy as do insulin-like growth factor-I, basic fibroblast growth factor (bFGF) and PDGF (10, 11). Although VEGF is known to be up-regulated primarily by hypoxia (12), some other factors (e.g. epidermal growth factor (13), prostaglandin EI and E2 (14), transforming growth factor-p (15, 16), interleukin-lp (17)) are also reported to induce VEGF production. Recently, the expression of VEGF mRNA was induced by estradiol in cell lines of rat uterine carcinoma, and its induction by estradiol, estriol or progesterone was also observed in the rat uterus in vivo (16, 18). However, it is not clear whether VEGF is up-regulated by estrogen or progesterone in cells derived from non-tumor, non-reproductive organs. To clarify whether VEGF is induced by a high concentration of estrogen or progesterone on retinal cells, and whether there are any synergistic effects between these hormones on VEGF secretion, we examined the VEGF production of cultured retinal cells under various concentrations of estrogen and progesterone.
Materials
and
methods
Cell Culture Primary cultures of bovine retinal pigment epithelial cells (RPE) were established as previously described with slight modifications (19). Briefly, the globes were bisected circumferentially 3 mm posterior to the limbus and the vitreous removed. The remaining eyecups were incubated with 0.25% trypsin in 0.02% EDTA (Gibco Laboratories, Grand Island, NY) for 15 min at 37”~. Removed RPE was transferred to collagen-coated flasks (Coming, New York, NY) and grown in Dulbecco’s modified Eagle’s medium (Gibco) supplemented with 10% fetal calf serum (Gibco) and 1% guinea pig serum (IBL, Gunma, Japan) under 5% CO2 at 37”~ and relative humidity of 95%. The media were changed every 3 days. At confluence, cells were subcultured using 0.25% trypsin in 0.02% EDTA (Gibco). Cells at the third passage were used for the experiment. Cells were identified histochemically by anti-cytokeratin antibody (Sigma Chemical Co., St. Louis, MO). VEGF production on high estrogen and/or progesterone concentrations RPE was re-plated in eleven plastic plates with six non-coated wells (Becton Dickinson Labware, Lincoln Park, NJ) and grown until confluency. The previously mentioned RPE medium but containing different concentrations of 17-OH estradiol p (Ez) (Sigma) (0.01, 1, 10, 100 FM), progesterone (Pd) (Sigma) (1, 2, 10,100 PM) and their combinations (E2 0.01 pM+Pd 1 PM, E2 1 ~M+PJ 10 PM), was prepared for each plate, while original RPE medium was supplied to a plate for the control group. As these agents are water-soluble products containing dextrin, dextrin was added to each well to make the amount of dextrin contained in each well equal. After incubation for 48 h, the conditioned media (1.5 ml per well) were collected and centrifuged. The supernatant was stored at -70 “C until assay. Determination of VEGF levels in cultured media The VEGF concentrations in the conditioned media were measured by our newly improved highly sensitive enzyme-linked immunosorbent assay (ELISA) using rabbit anti-human VEGF/VPF polyclonal antibody for both capture (solid phase) and secondary (enzyme-labeled) antibodies (20). This immunoassay detects all known isoforms of VEGF. The limit of detection for the assay, defined as +2SD above the zero standard, is 1 pg/ml. The intra-assay coefficient of variation was less than 6.1%. All the conditioned media were assayed after concentrating them to one-fourth of the original volume by vacuum centrifugation. Neither of these concentrating processes nor the concentrations of E 2, P4 or dextrin interfered with our VEGF assay (data not shown). All the measurements were corrected by protein content in each well measured by the method of Lowry et al. (21). All the results are expressed as mean&D and analysis of variance (ANOVA) was used for statistical analysis, and PcO.05 was taken as the level of significance.
VEGF Induction on Retina by Progesterone
Vol. 59, No. 1,1996
23
Results
In the Ez groups, only the VEGF concentrations at 100 PM of & (204.6k20.3 nglg protein, PcO.05) showed significant elevation, while the 0.01 PM (152.8+18.4 rig/g protein), 1 PM (171.7k22.2 rig/g protein) and 10 PM (153.5t23.7 rig/g protein) groups did not show significant changes compared to the control group (147.7k17.9 nglg protein) (Fig. 1). In the Pa group, 10 PM (214.51283 rig/g protein, PcO.01) and 100 PM (211.2&27.1 rig/g protein, P
E2b-W P4(lW
0 0 (control)
0.01 0
1 0
10 0
0 1
loo 0
:
0 10
0 loo
0.01 1
1 10
Fig. 1
Vascular endothelial growth factor levels in conditioned medium of bovine retinal pigment epithelial cells stimulated with various concentrations of estradiol and/or progesterone (N=6, Mean&D, * P&.0.5, ** PcO.01 compared with control group, NS; not significant)
Discussion
The reason for the worsening of diabetic retinopathy during pregnancy, especially pathological neovascularization or exudative changes frequently observed in patients with longterm diabetes, remains obscure. Too rapid normalization of blood glucose levels (22, 23) or hypertension associated with pregnancy (24) may be involved but these factors do not explain the entire phenomenon (12). Our findings demonstrated that the induction of VEGF by high concentrations of estradiol or progesterone is not limited to tumor cells and reproductive organs. Although the induction of VEGF by E2 at 0.01 PM was observed in HECl-A cells, a human endometrial adenocarcinoma cell line, it was seen only at concentrations far above the physiological range (100 PM) in RPE. This might be due to variability in the number of estrogen receptors that each cell possesses. (The distribution of estrogen receptors is not limited to the target tissues (25)) On the other hand, our findings indicated significant up-regulation of VEGF by physiologically elevated concentrations of P4 during pregnancy. During pregnancy, upregulation of P4, but not b, has been shown to contribute to the worsening of diabetic retinopathy through up-regulated VEGF. This hypothesis is supported by the clinical observations of Larinkari et al (5) who reported that, during the second trimester, significantly higher levels of serum progesterone but not estradiol were found in pregnant diabetic patients with retinopathy compared to those without retinopathy. The reasons why pathological retinal proliferation occurs
VEGF Inductionon Retina by Progesterone
24
Vol. 59, No. 1, 1996
only in diabetic but not in normal pregnant subjects may be the requirement of pre-existing background retinopathy and/or the need for hypoxia to induce sufficient up-regulation of VEGF. Although most reports determining stimulating factors of VEGF induction are based on elevation of its mRNA levels (12- 15), we confirmed such induction using the protein levels. Since our VEGF protein assay uses anti-human VEGF antibody, our measurements represent relative values, not the absolute values of the VEGF concentration in the conditioned medium. Four isoforms of VEGF are produced from a single gene by alternative splicing (7). The four isoforms each have 121, 165, 189,206 amino acids after removal of the signal peptide, and the former two are released from the producing cells, while the latter two bind to the extracellular matrix (26). Our assay using the antibody against the 121 amino acid form of VEGF detects all of those isoforms. Retinal pigment epithelium forms a single layer of cells which compose the outermost layer of the neural retina. Its contribution to diabetic retinopathy has been underestimated because of its location. However, the retinal pigment epithelium is known to form an integral part of the outer blood-retinal barrier, and its breakdown is known as the earliest pathological change in diabetic retinopathy (27). In a diabetic rat model, this breakdown was found to be due to alteration of membrane permeability rather than a loss of tight junctions (28). RPE-derived VEGF may play some roles in this process through autocrine pathways, because VEGF/VPF is known to mai kedly promote vascular permeability (7), as its name indicates. Moreover, RPE produces mitogenic factors for retinal microvascular cells (29) including bFGF (30), that has a synergistic effect with VEGF (3 1), and evidence that RPE contributes to the proliferative diabetic retinopathy (PDR) membrane has been obtained by ultrastructural investigation by Hamilton et al. (32). Furthermore, Hiscott et al. (33) histochemically showed in PDR patients that RPE is contained (520%) in combined traction rhegmatogenous retinal detachment membrane. These observations strongly suggest that RPE migrates through retinal breaks to access the PDR membrane and could contribute to its progression by secreting some angiogenic factors. In fact, the PDGF A chain was confirmed to be expressed in RPE of the PDR membranes (11). Half-maximal stimulation of endothelial cell growth is obtained at 100-150 pg/ml of VEGF (7), which is approximately the same concentration range of the RPE-conditioned media obtained in our study.
Acknowledement The authors thank Ms.Tomoe Matsumoto of Toagosei Co. Ltd. for her excellent technical assistance, and Chikusei Meat Center, All Japan Stockbreeding Public Corporation for their kind supply of bovine eyes. Thanks are also extended to Dr. Yasuo Sekine and Prof. Sachiko Honmura of the Department of Ophthalmology, University of Tsukuba for their encouragement.
References 1. 2.
B. KLEIN, S. MOSS and R. KLEIN, Diabetes Care. 12 34-40 (1990). K. ELMAN, R. WELCH, R. FRANK, G. GOYERT and R. SOKOL, Obstet Gynecol. 75
3.
L. JOVANOVIC-PETERSON and C. PETERSON, Clin Obstet Gynecol. 34 516-525 (1991). M. BOOTH and T. EL GARF, J Obstet Gynaecol Br Comwth. a 768-776 (1974). J. LARINKARI, L. LAATIKAINEN, T. RANTA, P. MORONEN, K. PESONEN and T. LAATIKAINEN, Diabetologia. 22 327-332 (1982). E. CORVAZIER, E. DUPUY, A. DOSNE and J. MACLOUF, Thromb Res. 34 303-310 (1984). N. FERRARA, K. HOUCK, L. JAKEMAN and D. LEUNG, Endocr Rev. 13 18-32 (1992). L. AIELLO, R. AVERY, P. ARIGG, B. KEYT, H. JAMPEL, S. SHAH, L. PASQUALE, H. THIEME, M. IWAMOTO, J. PARK, H. NGUYEN, L. AIELLO, N. FERRARA and G. KING, N Engl J Med. 331 1480-1487 (1994).
119-127 (1990).
4. 5. 6. 7. 8.
Vol. 59, No. 1, 1996
9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 2.5. 26. 27. 28. 29. 30. 31. 32. 33.
VEGF
Induction on Retina by Progesterone
A.P. ADAMIS, J.W. MILLER, M.T. BERNAL, D.J. D’AMICO, J. FOLKMAN, T.K. YE0 and K.T. YEO, Am J Ophthalmol. 118 445450 (1994). M. BARINAGA, Science. 267 452-453 (1995). S. ROBBINS, R. MIXON, D. WILSON, E. HART, J. ROBERTSON, I. WESTRA, S. PLANCK and J. ROSENBAUM, Invest Ophthalmol Vis Sci. 35 3649-3663 (1994). D. SHWEIKI, A. ITIN, D. SOFFER and E. KESHET, Nature. 359 843-845 (1992). C. GOLDMAN, J. KIM, W. WONG, V. KING, T. BROCK and G. GILLESPIE, Mol Biol Cell. 4 121-133 (1993). S. HARADA, J. NAGY, K. SULLIVAN, K. THOMAS, N. ENDO, G. RODAN and S. RODAN, J Clin Invest. 93 2490-24% (1994). L. PERTOVAARA, A. KAIPAINEN, T. MUSTONEN, A. ORPANA, N. FERRARA, 0. SAKSELA and K. ALITALO, J Biol Chem. 269 6271-6274 (1994). B. CULLINAN and R. KOOS, Endocrinology. 133 829-837 (1993). J. LI, M. PERRELLA, J. TSAI, S. YET, C. HSIEH, M. YOSHIZUMI, C. PATTERSON, W. ENDEGE, F. ZHOU and M. LEE, J Biol Chem. 270 308-3 12 (1995). J. CHARNOCK, A. SHARKEY, W. RAJPUT, D. BURCH, J. SCHOFIELD, S. FOUNTAIN, C. BOOCOCK and S. SMITH, Biol Reprod. 48 1120-l 128 (1993). A. ADAMIS, D. SHIMA, K. YEO, T. YEO, L. BROWN, B. BERSE, P. D’AMORE and J. FOLKMAN, Biochem Biophys Res Commun. 193 63 1-638 (1993). M. HANATANI, Y. TANAKA, S. KONDO, I. OMORI and H. SUZUKI, Biosci Biotech Biochem. 59 1958-1959 (1995). 0. LOWRY, N. ROSENBROUGH, A. FARR and J. RANDALL, J Biol Chem. &I 265 275 (1951). K. DAHL-JORGENSEN, 0. BRINCHMANN-HANSEN, K. HANSSEN, L. SANDVIK, F. AAGENSAES and A.D. GROU, Br Med J. 290 811-815 (1985). L. LAATIKAINEN, K. TERAMO, H.H. HIETA, V. KOIVISTO and R. PELKONEN, Acta Med Stand. 221 367-376 (1987). B. ROSENN, M. MIODOVNIK, G. KRANIAS, J. KHOURY, C. COMBS, F. MIMOUNI, T. SIDDIQI and M. LIPMAM, Am J Obstet Gynecol. 166 1214-1218 (1992). D. CIOCCA and L. ROIG, Endcr Rev. 16 35-62 (1995). K. HOUCK, D. LEUNG, A. ROWLAND, J. WINER and N. FERRARA, J Biol Chem. 267 26031-26037 (1992). S. WALTAM, C. OESTRICH, T. KRUPIN, S. HANISH, S. PATZAN, J. SANTIAGO and C. KILO, Diabetes. 27 85-87 (1978). R. CALDWELL, S. SLAPNICK and B. MCLAUGHLIN, Curr Eye Res. 4 215-227 (1985). H. WONG, M. BOULTON, P. CLARK, M. BAYLY and J. MARSHALL, Eye. 1. 754756 (1987). L. BOST, A. AOTAKI-KEEN and L. HJELMELAND, Exp Eye Res. 55 727-734 (1992). F. GOTO, K. GOTO, K. WEINDEL and J. FOLKMAN, Lab Invest. @ 508-517 (1993). C. HAMILTON, D. CHANDLER, G. KLINTWORTH and R. MACHMER, Am J Ophthalmol. 94 479-488 (1982). P. HISCOTT, R. GRAY, I. GRIERSON and Z. GREGOR, Br J Ophthalmol. 2s 219-222 (1994).
25