|Molecular Vision 1999;
Received 15 July 1999 | Accepted 28 December 1999 | Published 29 December 1999
The mRNA phenotype of a human RPE cell line at replicative senescence
Hiroshi Matsunaga,1 James T. Handa,1 Claire Mazow
Gelfman,2 Leonard M.
Departments of 1Ophthalmology and 2Molecular and Cellular Biology, University of California-Davis, Davis, CA
Correspondence to: Leonard M. Hjelmeland, PhD, Vitreoretinal Research Laboratory, School of Medicine, University of California, One Shields Avenue, Davis, CA, 95616-8794; Phone: (530) 752-2250; FAX: (530) 752-2270; email: firstname.lastname@example.org
Purpose: To explore the changes in expression of a set of genes in a single retinal pigment epithelial (RPE) cell line and two fibroblast cell lines as controls under culture conditions previously used for the analysis of senescent gene expression.
Methods: A single human RPE cell line, which had previously been characterized using known markers of senescence, and two fibroblast cell lines were grown to replicative exhaustion. The mRNA phenotype of genes known to be altered by senescence were studied by quantitative Northern analysis.
Results: The mRNA phenotype of cells changes at replicative senescence yielding a synthetic phenotype which is similar to cells found in repairing wounds. Of the genes studied, urokinase-type plasminogen activator and plasminogen activator inhibitor-1 were regulated in RPE cells similar to fibroblasts at senescence. The largest changes noted for any single gene were the upregulation of insulin growth factor binding protein 2, and the downregulation of collagen I alpha 2, basic fibroblast growth factor, and fibroblast growth factor-5.
Conclusions: This study demonstrates an altered mRNA phenotype of a human RPE cell line grown to replicative exhaustion. This analysis of a single cell line emphasizes the variability of results based on a single cell line or tissue specimen and indicates the need for additional study.
The finite replicative capacity of human cells in vitro has been widely observed and studied since the original publications of Hayflick and others [1-4]. At the end of the replicative lifespan in culture, these cells display an altered morphology and are unable to reenter the cell cycle despite the addition of growth factors. This specific state of replicative exhaustion is now known as replicative senescence, and is clearly distinguishable from quiescence based on the ability of quiescent cells to reenter the cell cycle. Several theories have been proposed to explain the mechanism controlling replicative senescence of human cells in culture. It appears that critical reduction in the length of chromosomal telomeres is sufficient to induce replicative senescence and the maintenance of adequate telomere length through the expression of the human telomere reverse transcriptase subunit of telomerase prevents replicative senescence [5,6].
The aging of human retinal pigment epithelial (RPE) cells in vitro has also been the subject of several studies. Flood et al. observed that the growth rate of RPE primary cultures as well as the number of large flattened nondividing cells was directly related to the chronological age of the donor . Burke and several coworkers systematically studied the properties of bovine and human RPE cultures that were intentionally aged through repeated passage [8-10]. Multiply passaged cultures exhibited lower growth rates, lower saturation densities, and an accumulation of cells which had a markedly altered morphology. These findings are very similar to observations made on a variety of senescent human cell cultures. Interestingly, Burke also showed that RPE cells from the macula of the human eye or the area centralis of the bovine eye had more limited replicative life-spans in vitro than did cells from more peripheral regions of the posterior pole. Recently, our laboratory showed that human RPE cells at replicative exhaustion have critically shortened telomeres and also express senescence-associated beta galactosidase activity, a convenient marker for the replicative senescence of human cells in vitro [11,12].
A serious attempt has been made to describe the molecular phenotype of senescent cells. Subtractive hybridization  and enhanced differential display  approaches were used to catalogue changes in gene expression for human fibroblasts at replicative senescence. The results of these surveys indicate that in addition to cell cycle regulated genes, a large number of the mRNAs that were differentially regulated in cellular senescence belonged to the classes of extracellular matrix proteins, extracellular proteases and their inhibitors, and some extracellular growth regulators. Two studies have appeared that are relevant to the replicative senescence of RPE cells. Guillonneau et al. showed that perlecan expression is upregulated with serial passage of RPE cells in vitro  and Tombran-Tink et al. suggested that pigment epithelium-derived factor (PEDF) is down regulated as a function of cell passage . Along with its role as a trophic factor, PEDF also belongs to a family of structurally related serine protease inhibitors. PEDF was independently cloned as Early Passage cDNA-1 (EPC-1) in a study of senescent gene expression in the WI-38 human fibroblast cell line .
To date, few studies have investigated the presence of senescent cells in vivo or the possible relevance of senescent cells to specific pathologies. It is difficult to imagine for example, how replicative senescence of RPE cells would be relevant for the RPE in vivo. Adult RPE cells constitute a nondividing population under nonpathological conditions and should not be subject to replicative exhaustion. We have recently presented an hypothesis concerning a mechanism through which RPE cells may become senescent in vivo. The proposed mechanism depends heavily on oxidative damage to chromosomal telomeres and effectively reduces or eliminates the requirement for large numbers of cell divisions in order for RPE cells to become senescent . Since the phenotype of replicatively senescent cells is related to alterations in the synthesis and turnover of extracellular matrix and altered levels of cellular growth regulators, it seems reasonable to characterize in depth the molecular phenotype of replicatively senescent RPE cells in terms of the number of genes studied and under conditions that allow useful comparisons with the extensive literature on senescent gene expression in human fibroblastic cells. It is tempting for example, to speculate that the reported downregulation of PEDF expression that occurs in serially passaged RPE cells may have some relationship to the function of any senescent RPE cells that might be present in human eyes with age-related macular degeneration. Although studies of human pathology have not yet appeared in the literature, our laboratory has recently published a method for detecting senescence associated ß-galactosidase activity in vivo . This study showed that senescence associated ß-galactosidase positive RPE cells appeared in the macula in aging rhesus eyes and were conspicuously present adjacent to cuticular drusen.
The purpose of this study was to examine an expanded set of candidate genes at replicative senescence by semi-quantitative Northern analysis under the identical culture conditions used for analysis of senescent gene expression used in senescent fibroblasts. The genes chosen to study include extracellular matrix proteins, proteases, protease inhibitors, and several growth factors in a single RPE cell line that had previously been characterized with respect to replicative senescence . Two reference nontransformed human fibroblast lines were also studied to make direct comparisons possible.
We previously described the serial passage of a single human RPE cell line from a 1 year old male trauma victim (RPE 340) to replicative senescence and the concurrent measurement of 5-bromo-2'-deoxyuridine labeling indices and senescence-associated beta galactosidase staining . For this study, we used the identical cell line at population doubling (PDL) 25 (young) and 52 (senescent). Briefly, cells were serially passaged by trypsinization at a split ratio of 1:4 and the population doublings for each passage were estimated assuming 2 PDLs per passage. The initial plating of the primary culture was assumed to be PDL 0. Cultures were maintained in Dulbecco's Modified Eagle medium/Nutrient mixture F12 with 15 mM Hepes buffer (DMEM/F12; Gibco, Inc., Grand Island, NY) + 10% fetal bovine serum (FBS; UBI Upstate, Lake Placid, NY), 0.348% additional sodium bicarbonate, 2 mM L-glutamine solution, (Gibco, Inc.) at 37 °C in 10% CO2 as previously described .
For serum withdrawal studies, sparse cultures were placed in DMEM/F12 + 0.1% FBS for 3 days, fed with fresh 0.1% serum containing media and harvested after an additional 2 days. Parallel cultures in DMEM/F12 + 10% FBS were maintained in this medium for 3 days and then refed with fresh 10% serum containing media and harvested after an additional 2 days. Cultures were grown for 3 weeks to obtain confluence by visual inspection before being subjected to the serum conditions outlined above for sparse cultures. For comparison, control studies of the human dermal B/J at young (PDL 25) and senescence (PDL 54) and lung IMR-90 fibroblast cell lines at young (PDL 25) and senescence (PDL 54) at subconfluence were performed, as previously described .
Complementary DNA Probes for Northern Analysis
The following cDNA probes were used for Northern analysis. A 931 bp EcoR1 perlecan cDNA fragment was provided by Dr. R.V. Iozzo . Dr. S. Hackett provided a 1.7 kb EcoR1/Bgl II fragment containing plasminogen activator inhibitor-1 (PAI-1) cDNA isolated from pPAI17  and a 1.5 kb Pst1 fragment containing urokinase-type plasminogen activator (u-PA) cDNA . The probe for pigment epithelium-derived factor (PEDF), from Dr. S.P. Becerra, was a 1.5 kb EcoR1/Hind III fragment isolated from pFS17 . Dr. D. Bok kindly provided the probe for tissue inhibitor of metalloproteinase-3 (TIMP-3), which was a 1240 bp EcoR1/Xhol fragment . The cDNA probe for Collagen I alpha 2 is a 1.3 kb NOT 1/EcoR1 fragment provided by Dr. W. Funk.
The following probes for the growth factors of interest were used in this study. The basic fibroblast growth factor (bFGF) cDNA probe was a 0.8 kb EcoR 1 fragment from plasmid pHFL1-7 . The fibroblast growth factor 5 (FGF-5) cDNA probe was a 1.0 kb EcoR1 fragment from pLTR122E . A 2.7 kb EcoR I fragment containing platelet derived growth factor-B (PDGF-B) cDNA was isolated as described by Johnsson et al. . The transforming growth factor-ß2 (TGF-ß2) probe is a 2.37 kb cDNA BamH1/Pst1 fragment from pcD-GIG2, and was a gift from Dr. R. Armour . The vascular endothelial growth factor (VEGF) cDNA probe was a 930 bp EcoR1 fragment kindly provided by Dr. A. Singh (Genentech, Inc., South San Francisco, CA). The insulin-like growth factor binding protein-2 (IGFBP-2) cDNA probe was a 446 bp EcoR I/HindIII fragment provided by Dr. S. Shimasaki . The 28S rRNA cDNA probe was a 7 kb EcoR1 fragment as described by Gonzalez et al. .
Northern Blot Analysis
Northern blot analysis for all samples was performed with random primed 32P labeled cDNA probes. Using the guanidium thiocyanate technique, total RNA was isolated from early and senescent cultures at both sparse and confluent densities in medium containing either 0.1 or 10% fetal bovine serum. Total RNA (10 µg) was subjected to electrophoresis through a 1.2% agarose gel in 6.6% formaldehyde, transferred to a nylon membrane and cross-linked using a Stratalinker (Stratagene, La Jolla, CA). Blots were then prehybridized for 2 h at 42 °C in 50% formamide, 5X SSC, 0.1% sodium dodecyl sulfate (SDS), 5X Denhardt's solution, and 100 µg denatured salmon sperm DNA. Hybridization was initiated by addition of denatured probe and continued at 42 °C for 15 h. Blots were then washed twice in 0.1X SSC / 0.1% SDS at room temperature for 5 min, and in 0.1X / 0.1% SDS at 50 °C for 2 h. Values were normalized using a 28S rRNA probe. Experiments were repeated at least twice.
All the membranes were placed onto a phosphorimager screen to quantify the signal intensity using the Molecular Analyst software (Bio Rad, Richmond, CA). A rectangle was drawn around each band, and the mean count of the individual bands was determined. The background was determined by averaging the counts of two equivalent rectangles above and below each band, which was subtracted from the mean count (Figure 1). All bands were normalized among lanes by reprobing with a 28S rRNA probe and quantifying the signal. The ratio of senescent to young RPE cell gene expression was calculated by dividing the values obtained for senescent and young cells in each condition.
The rationale for our initial studies was to choose a small set of genes for comparison of replicative senescence between cultures of human RPE cells and human fibroblasts. Criteria for the selection of genes included expression in the RPE and suggested or demonstrated alteration of expression with aging or at replicative senescence. Several genes in the categories of matrix proteins, extracellular proteases, and protease inhibitors met these criteria. PEDF and perlecan were suggested to have altered levels of expression as a function of passage number in RPE cells [15,16], while the altered expression of PAI-1, u-PA, and TIMP-3 have been demonstrated in a variety of fibroblast lines grown to senescence [32,33]. The relative ratio of the expression of each of these genes between senescent and young cultures of RPE cells was quantified by Northern blot analysis in both 0.1% and 10% serum. These data are presented in Table 1 and Figure 2. Normalization for these Northern blots were conducted by 28S rRNA hybridization and are presented in Figure 3. To facilitate a comparison with reported results in the literature, the plating density (subconfluence) and serum conditions (0.1% and 10% serum) used in this study were identical to those utilized in previously published studies on the replicative senescence of fibroblasts . The original rationale for the design of these studies was based on the need to distinguish between gene expression events related to withdrawal from the cell cycle into the state of replicative senescence as opposed to events related to withdrawal into quiescence. Senescence and quiescence can be distinguished by the ability of quiescent cells to reenter the cell cycle when challenged with serum. The original observations of cells entering the senescent state were made in sparse cultures in 10% serum. The addition, therefore, of parallel cultures in 0.1% serum provides experimental conditions where gene expression can be compared between quiescent and senescent cells. Our control studies for the B/J and IMR-90 fibroblast lines are also reported in Table 1.
The assumption that 0.1% serum treatment causes efficient withdrawal from the cell cycle for RPE cells may not be valid. As a result, cells were grown to a postconfluent state and the analysis was repeated. The results are presented in Table 2 and Figure 2. Control studies (data not shown) for these cultures indicated a very low labeling index, defined as the percentage of cells that incorporate BrdU into the nuclear DNA for 72 h , indicating that cells were truly quiescent, which allows a better comparison with senescent cultures.
Growth factors represent another class of proteins of interest to our laboratory whose expression might be altered during senescence of RPE cells. We assembled a short list of factors that may be involved in angiogenesis and fibrosis in the posterior pole during aging. Once again, these genes were studied at sparse (Table 3) and confluent densities (Table 4) with the procedures described above. Under all conditions, bFGF had the most dramatically altered expression in RPE cultures at senescence compared to young PDL of all factors surveyed (Figure 4). FGF-5 was downregulated in both sparse and confluent RPE cultures grown in low serum.
During our experiments involving Northern analysis, the same mRNA populations were examined by microarray analysis (Dr. W. Funk, personal communication, 1999). This analysis is the subject of a separate publication. The genes whose expression was most dramatically altered at senescence fell into the categories of growth regulators and matrix proteins. IGFBP-2 was one of the genes found to be most dramatically upregulated in replicative senescence for a variety of cell lines. We explored IGFBP-2 synthesis in sparse and confluent RPE cells by Northern analysis and confirmed this finding (Table 3, Table 4, and Figure 5). The microarray results indicated that collagen I [alpha]2 expression was dramatically downregulated in confluent senescent RPE cells, and we also confirmed this finding by Northern blot analysis (Figure 2F). In contrast, collagen I [alpha]2 expression was upregulated in subconfluent RPE cells grown in high serum (Table 1, Table 2, and Figure 2F).
Using a range of <=0.5 and >=2.0 in the ratio of gene expression between senescent and young cultures as an initial filter, only a few of the genes studied by Northern analyses appeared to be significant. Of these, u-PA and PAI-1 have been previously discussed in the fibroblast literature with respect to replicative senescence , and perlecan was also suggested to be upregulated with serial passage of the RPE . u-PA was upregulated in senescent RPE cells only under high serum conditions. We speculate that a differential response to serum by senescent as opposed to young RPE cells resulted in increased u-PA expression. Interestingly, bFGF and FGF-5 were significantly downregulated at replicative senescence in many of the culture conditions we studied as quantified by Northern analysis. The dramatic alterations of IGFBP-2 and collagen I [alpha]2 expression confirm the microarray results reported elsewhere.
We could not confirm the downregulation of PEDF/EPC-1 and TIMP-3 that were previously reported [16,17,33]. A direct comparison between the B/J and IMR-90 human fibroblast cell lines in our study revealed the difficulty of making generalizations based on single cell lines, although B/J is a dermal line and IMR-90 is derived from lung. This issue is obviously relevant to the current study where data from only one cell line are reported. Convincing generalizations can be made only after examination of multiple RPE cell lines.
Our Northern analysis disagreed with previous reports concerning PEDF/EPC-1 and TIMP-3 downregulation during replicative senescence [16,17,33]. Tombran-Tink et al. investigated PEDF expression in confluent monolayers of serially passaged monkey RPE cells and observed downregulation of the expression of the PEDF mRNA by Northern analysis and protein by Western blot analysis . These results may vary from our own based on the difference in species, culture conditions, differences between the analysis of mRNA and protein expression, or more likely the variability of individual lines that are propagated in vitro. Pignolo et al. independently cloned the same gene by subtractive hybridization of RNAs expressed at early and senescent passages of the WI-38 human fibroblast cell line . PEDF/EPC-1 was selected based on the dramatic downregulation at senescence. Not only did we not observe such a downregulation in RPE cells, but we also could not observe downregulation in either the B/J or IMR-90 cell lines. Once again, it is probable that differences among individual lines propagated in vitro are real and thus generalizations should be made with a great deal of caution. It is interesting to ask if these differences represent genetic differences among donors, ex vivo variation as a function of serial propagation in culture, or even the presence of multiple phenotypes with respect to the regulation of individual genes within an otherwise apparently homogenous cell population like the RPE in vivo.
TIMP-3 represents the second literature observation we could not reproduce. TIMP-3 was originally identified as Mig-5, and its expression and downregulation as a function of serial passage in the WI-38 line was studied . Once again, we could not confirm this observation in any of the three lines we studied, and the explanation for this observation is likely to be the same given for our observations on PEDF/EPC-1. Previous reports in the fibroblast literature have attempted to define a phenotype for replicatively senescent cells based on alterations in gene expression [13,14]. Essentially, genes for matrix proteins, proteases, protease inhibitors, and inflammatory cytokines are altered in a fashion that would make the senescent phenotype similar to a wound repair phenotype for fibroblasts. It is premature to draw any such conclusions from our study, primarily due to the fact that a single RPE cell line was examined, but the results from this study encourage us to investigate in more detail the altered gene expression by senescent RPE cells. It will be especially interesting to see if the expression of bFGF, u-PA, PAI-1, IGFBP-2, and collagen I [alpha]2 are similarly regulated at senescence in multiple, independent RPE lines.
One of the unexpected findings in our study was the altered gene expression of bFGF at senescence. To our knowledge, no previous report has documented the altered expression of bFGF in replicative senescence. Basic FGF has been shown to be a survival factor for photoreceptors, RPE cells, and endothelial cells [34-38]. Because of its potential autocrine and paracrine trophic role in the outer retina, downregulation of this gene at senescence could adversely impact the viability of many different cell types in the vicinity of the RPE.
The larger question for this work is how or if the study of replicative senescence in vitro has any relevance to the RPE in vivo. Clearly, RPE cells in vivo are nondividing except in circumstances where they are actively involved in a wound repair process. It will be necessary to critically examine the aging human eye for the presence of senescent RPE cells using markers for senescence that have been unambiguously identified for multiple RPE lines using the approaches outlined in the current study. If these analyses generalize our current observation on the expression of IGFBP-2, this gene might serve as a useful marker of senescence in vivo. Second, it will also be necessary to explore mechanisms other than simple replicative exhaustion which might explain how nondividing cells could enter the senescence program. The recent studies of von Zylinicki et al. provide an attractive insight into the possible role of oxidative stress in creating persistent damage to chromosomal telomeres [39-41]. Since shortening of telomeres has now been demonstrated to be sufficient to cause cells to enter senescence and the maintenance of telomere length by telomerase will prevent this process in RPE cells , these observations may explain a plausible pathway leading to senescence of the RPE in vivo. We have recently published an editorial that proposes a hypothesis for a mechanism leading specifically to photo-oxidative damage of telomeres in the RPE cells of the macula . Ultimately, if senescent RPE cells can be carefully documented in vivo, and if such cells exhibit accumulation with age and a primary distribution in the macula, some relationship may exist between senescence of the RPE and the etiology of age-related macular degeneration.
This work was supported by National Eye Institute grants EY06473 (LMH), EY00344 (JTH), and EY 06650 (CMG); Research to Prevent Blindness Senior Scientist Award (LMH); Research to Prevent Blindness Manpower Award (JTH); and an unrestricted grant from Research to Prevent Blindness.
1. Hayflick L, Moorhead P. The serial cultivation of human diploid cell strains. Exp Cell Res 1961; 25:585-621.
2. Hayflick L. The limited in vitro lifetime of human diploid cell strains. Exp Cell Res 1965; 37:614-36.
3. Martin GM, Sprague CA, Epstein CJ. Replicative life-span of cultivated human cells. Effects of donor's age, tissue, and genotype. Lab Invest 1970; 23:86-92.
4. Schneider EL, Mitsui Y. The relationship between in vitro cellular aging and in vivo human age. Proc Natl Acad Sci U S A 1976; 73:3584-8.
5. West MD. The cellular and molecular biology of skin aging. Arch Dermatol 1994; 130:87-95.
6. Bodnar AG, Ouellette M, Frolkis M, Holt SE, Chiu CP, Morin GB, Harley CB, Shay JW, Lichtsteiner S, Wright WE. Extension of life-span by introduction of telomerase into normal human cells. Science 1998; 279:349-52.
7. Flood MT, Gouras P, Kjeldbye H. Growth characteristics and ultrastructure of human retinal pigment epithelium in vitro. Invest Ophthalmol Vis Sci 1980; 19:1309-20.
8. Burke JM, Soref C. Topographical variation in growth in cultured bovine retinal pigment epithelium. Invest Ophthalmol Vis Sci 1988; 29:1784-8.
9. Burke JM, McKay BS. In vitro aging of bovine and human retinal pigment epithelium: number and activity of the Na/K ATPase pump. Exp Eye Res 1993; 57:51-7.
10. Burke JM. Cytochrome oxidase activity in bovine and human retinal pigment epithelium: topographical and age-related differences. Curr Eye Res 1993; 12:1073-9.
11. Dimri GP, Lee X, Basile G, Acosta M, Scott G, Roskelley C, Medrano EE, Linskens M, Rubelj I, Pereira-Smith O, et al. A biomarker that identifies senescent human cells in culture and in aging skin in vivo. Proc Natl Acad Sci U S A 1995; 92:9363-7.
12. Matsunaga H, Handa JT, Aotaki-Keen A, Sherwood SW, West MD, Hjelmeland LM. Beta-galactosidase histochemistry and telomere loss in senescent retinal pigment epithelial cells. Invest Ophthalmol Vis Sci 1999; 40:197-202.
13. Cristofalo VJ, Pignolo RJ. Molecular markers of senescence in fibroblast-like cultures. Exp Gerontol 1996; 31:111-23.
14. Linskens MH, Feng J, Andrews WH, Enlow BE, Saati SM, Tonkin LA, Funk WD, Villeponteau B. Cataloging altered gene expression in young and senescent cells using enhanced differential display. Nucleic Acids Res 1995; 23:3244-51.
15. Guillonneau X, Tassin J, Berrou E, Bryckaert M, Courtois Y, Mascarelli F. In vitro changes in plasma membrane heparan sulfate proteoglycans and in perlecan expression participate in the regulation of fibroblast growth factor 2 mitogenic activity. J Cell Physiol 1996; 166:170-87.
16. Tombran-Tink J, Shivaram SM, Chader GJ, Johnson LV, Bok D. Expression, secretion, and age-related downregulation of pigment epithelium-derived factor, a serpin with neurotrophic activity. J Neurosci 1995; 15:4992-5003.
17. Pignolo RJ, Rotenberg MO, Cristofalo VJ. Analysis of EPC-1 growth state-dependent expression, specificity, and conservation of related sequences. J Cell Physiol 1995; 162:110-8.
18. Hjelmeland LM. Senescence of the retinal pigmented epithelium. Invest Ophthalmol Vis Sci 1999; 40:1-2.
19. Mishima K, Handa JT, Aotaki-Keen A, Lutty GA, Morse LS, Hjelmeland LM. Senescence-associated beta-galactosidase histochemistry for the primate eye. Invest Ophthalmol Vis Sci 1999; 40:1590-3.
20. Ishigooka H, Aotaki-Keen AE, Hjelmeland LM. Subcellular localization of bFGF in human retinal pigment epithelium in vitro. Exp Eye Res 1992; 55:203-14.
21. Murdoch AD, Dodge GR, Cohen I, Tuan RS, Iozzo RV. Primary structure of the human heparan sulfate proteoglycan from basement membrane (HSPG2/perlecan). A chimeric molecule with multiple domains homologous to the low density lipoprotein receptor, laminin, neural cell adhesion molecules, and epidermal growth factor. J Biol Chem 1992; 267:8544-57.
22. Ginsburg D, Zeheb R, Yang AY, Rafferty UM, Andreasen PA, Nielsen L, Dano K, Lebo RV, Gelehrter TD. cDNA cloning of human plasminogen activator-inhibitor from endothelial cells. J Clin Invest 1986; 78:1673-80.
23. Verde P, Stoppelli MP, Galeffi P, Di Nocera P, Blasi F. Identification and primary sequence of an unspliced human urokinase poly(A)+ RNA. Proc Natl Acad Sci U S A 1984; 81:4727-31.
24. Steele FR, Chader GJ, Johnson LV, Tombran-Tink J. Pigment epithelium-derived factor: neurotrophic activity and identification as a member of the serine protease inhibitor gene family. Proc Natl Acad Sci U S A 1993; 90:1526-30.
25. Ruiz A, Brett P, Bok D. TIMP-3 is expressed in the human retinal pigment epithelium. Biochem Biophys Res Commun 1996; 226:467-74.
26. Abraham JA, Whang JL, Tumolo A, Mergia A, Friedman J, Gospodarowicz D, Fiddes JC. Human basic fibroblast growth factor: nucleotide sequence and genomic organization. EMBO J 1986; 5:2523-8.
27. Zhan X, Bates B, Hu XG, Goldfarb M. The human FGF-5 oncogene encodes a novel protein related to fibroblast growth factors. Mol Cell Biol 1988; 8:3487-95.
28. Johnsson A, Heldin CH, Wasteson A, Westermark B, Deuel TF, Huang JS, Seeburg PH, Gray A, Ullrich A, Scrace G, et al. The c-sis gene encodes a recursor of the B chain of platelet-derived growth factor. EMBO J 1984; 3:921-8.
29. Hanks SK, Armour R, Baldwin JH, Maldonado F, Spiess J, Holley RW. Amino acid sequence of the BSC-1 cell growth inhibitor (polyergin) deduced from the nucleotide sequence of the cDNA. Proc Natl Acad Sci U S A 1988; 85:79-82.
30. Binkert C, Landwehr J, Mary JL, Schwander J, Heinrich G. Cloning, sequence analysis and expression of a cDNA encoding a novel insulin-like growth factor binding protein (IGFBP-2). EMBO J 1989; 8:2497-502.
31. Gonzalez IL, Gorski JL, Campen TJ, Dorney DJ, Erickson JM, Sylvester JE, Schmickel RD. Variation among human 28S ribosomal RNA genes. Proc Natl Acad Sci U S A 1985; 82:7666-70.
32. West MD, Shay JW, Wright WE, Linskens MH. Altered expression of plasminogen activator and plasminogen activator inhibitor during cellular senescence. Exp Gerontol 1996; 31:175-93.
33. Wick M, Burger C, Brusselbach S, Lucibello FC, Muller R. A novel member of human tissue inhibitor of metalloproteinases (TIMP) gene family is regulated during G1 progression, mitogenic stimulation, differentiation, and senescence. J Biol Chem 1994; 269:18953-60.
34. Dreyfus H, Sahel J, Heidinger V, Mohand-Said S, Guerold B, Meuillet E, Fontaine V, Hicks D. Gangliosides and neurotrophic growth factors in the retina. Molecular interactions and applications as neuroprotective agents. Ann N Y Acad Sci 1998; 845:240-52.
35. Faktorovich EG, Steinberg RH, Yasumura D, Matthes MT, LaVail MM. Photoreceptor degeneration in inherited retinal dystrophy delayed by basic fibroblast growth factor. Nature 1990; 347:83-6.
36. Faktorovich EG, Steinberg RH, Yasumura D, Matthes MT, LaVail MM. Basic fibroblast growth factor and local injury protect photoreceptors from light damage in the rat. J Neurosci 1992; 12:3554-67.
37. Hicks D, Courtois Y. Fibroblast growth factor stimulates photoreceptor differentiation in vitro. J Neurosci 1992; 12:2022-33.
38. Speliotes EK, Caday CG, Do T, Weise J, Kowall NW, Finklestein SP. Increased expression of basic fibroblast growth factor (bFGF) following focal cerebral infarction in the rat. Brain Res Mol Brain Res 1996; 39:31-42.
39. von Zglinicki T, Saretzki G, Docke W, Lotze C. Mild hyperoxia shortens telomeres and inhibits proliferation of fibroblasts: a model for senescence? Exp Cell Res 1995; 220:186-93.
40. Sitte N, Saretzki G, von Zglinicki T. Accelerated telomere shortening in fibroblasts after extended periods of confluency. Free Radic Biol Med 1998; 24:885-93.
41. Petersen S, Saretzki G, von Zglinicki T. Preferential accumulation of single-stranded regions in telomeres of human fibroblasts. Exp Cell Res 1998; 239:152-60.