The influence of the genotype on the process of ageing of chick lens cells in vitro

Experimental Cell Research 174 (1988) 330-343

The Influence

of the Genotype on the Process of Ageing of Chick Lens Cells in Vitro C. E. PATEK’ and R. M. CLAYTON*

Department of Genetics, University of Edinburgh, King’s Buildings, West Mains Road, Edinburgh EH9 3JN, United Kingdom

We reported previously that changes in crystallin expression in differentiating long-term primary cultures of lens cells from five different chick genotypes are similar to those which occur in uiuo between hatching and the &week-old adult. These changes followed a similar program in all genotypes but occurred more rapidly in cells from the fast-growing than from the slow-growing genotypes. The present study examines ageing changes in lens cell populations from the same five genotypes, over a 4-6 month period, using long-term serial subcultures. The capacity for lentoid differentiation was progressively lost, but the rate of loss was inversely related to the intrinsic growth rate of the cells of these genotypes, occurring at the first passage in the slowest-growing strain, while fifth passage cells of the fastest-growing strain still retained some lentoid-forming capacity. The rate of loss of crystallin expression was also inversely related to the genetic growth rate, but the sequence of changes appears to be nonrandom, since it was broadly similar in all genotypes, starting with a preferential loss of Scrystallin, as occurs in uivo; although a- and Bcrystallins were undetectable in late dedifferentiated cultures, the capacity of the cells for their synthesis was still present. Cultures from both fast-growing genotypes eventually showed senescence, but those from all three slow-growing genotypes underwent transformation. The major cell component in late cultures of all genotypes was actin. 0 1988 Academic

Press, Inc.

The growth of the vertebrate lens and the differentiation of the epithelial cells into lens tiber cells continues throughout life, but the representation of the crystallin polypeptides changes during development and ageing so that lens fiber cells have a crystallin composition related to their position in the lens and the age of the animal when they are laid down (reviewed by [l, 21). Fiber cells can differentiate in vitro forming lentoid bodies, which contain high levels of crystallins [3-8]. The sequence of changes in crystallin and membrane protein expression which occurs in long-term primary cultures of chick lens epithelial cells is similar to those which occur during development in uiuo [9, 101, and were found to be similar in five different genotypes distinguished by their growth rate. The rate of change was directly related to the genotype of the cells, being most rapid in the cells growing fastest in vitro and slowest in the cells growing most slowly in vitro [ll]. Of these five unrelated genotypes, Hy-1 and Hy-2 are fast-growing, show some degree of lens epithelial hyperplasia in uiuo, and anomalies of the cell surface, and of cell behavior in vitro [12-161. N-Rd, N-Db and N-J are slowgrowing strains with normal lens morphology in uiuo. A comparison of the ’ Present address: Department of Pathology, University of Edinburgh Medical School, Teviot Place, Edinburgh EH8 9AG. ’ To whom reprint requests should be addressed. Copyright @I 1988 by Academic Press, Inc. AU rights of reproduction in any form reserved 0014-4827/88 $03.00

330

Lens cell ageing

331

molecular changes during cellular ageing in vitro of these same five genotypes and a comparison with changes in the ageing lens in uiuo permit us to examine the nature of any genetic contribution to the process of cellular ageing. To this end, we have maintained cell populations derived from day-old chick lens epithelial cells of these genotypes for up to 6 months by serial subculture and examined the changes in lentoid-forming capacity and crystallin content. Since the lens is enclosed in an acelluar capsule and contains only the lens epithelial cells and the lens fibers into which they terminally differentiate, pure cell cultures are readily established. Since the metabolic integrity as well as the growth of the lens depends upon the epithelial cells, age-related changes may also contribute to cataractogenesis [see 2, 17, 181.

MATERIALS

AND METHODS

Lens epithelial cells (LEC) from day-old chicks of the five genotypes were rolled on sterile filter paper to remove any adherent fragments of iris. The lenses were placed in calcium-magnesium-free saline (CMF) and punctured at the posterior pole, to remove the fiber cell mass. The epithelial cells were sucked out by pipet from within the capsule, dissociated, washed, and plated out in Eagle’s MEM with 6% fetal calf serum as described previously [3-91. This procedure ensures the establishment of pure lens epithelial cultures, uncontaminated by any other cell type. Two separate sets of cultures were established at different times for each genotype and both were carried throughout the serial passage process and investigated as described below. One batch of fetal calf serum (GIBCO Biocult, Ltd.) was used throughout the study to avoid any variation in growth conditions. Differentiated LEC cultures were serially cultured at 28-day intervals, and each time the epithelial cells were replated at the original cell density. Cells were labeled with a ‘H amino acid mixture (20-50 uCi/ml/2-5 h; Amersham International, Amersham, England) and harvested and the cell number was estimated as described previously [9]. Proteins were extracted from lens cultures, and also from decapsulated day-old and I-week-old adult chick fiber masses, and analyzed by SDS-polyacrylamide gel electrophoresis (SDS-PAGE) and fluorography as previously described [9].

RESULTS Cell Growth and Morphology

The series of morphological changes during the early stages of differentiation of the primary cultures with the appearance of groups of polygonal cells, and later, of lentoids, in the originally homogeneous epithelial cell sheet (Figs. 1a-c) were as described previously [3-131. The plating efficiency of the three slow-growing strains was from 70 to 80%, but was 3550% for the fast-growing strain. Growth during log phase was more rapid in Hy-1 and Hy-2 than in the other strains (Fig. 2) but by Day 28, all the strains, except for the slowest-growing N-Rd, had achieved a similar cell density. Lentoid bodies appeared about 3 days earlier and were larger and more numerous in the fast-growing strains. During the process of serial subculture we found that the rate of loss of the capacity for lentoid body differentiation was inversely related to the initial growth rate of these genotypes. No lentoid bodies appeared in N-Rd cultures after their first reseeding (that is, second passage cultures), in third passage N-J and N-Db cultures, nor in the fourth passage of Hy-2 cultures (that is, after the third reseeding). Lentoid

332 Patek and Clayton

Fig. 1. Photomicrographs (x300) of LEC at different stages of cell passage. Primary LEC showing (a) undifferentiated epithelial cells, (b) “islets” of small polygonal-shaped cells, and (c) lentoid bodies with bottle cells. Passaged cultures showing (d) small lentoid bodies associated with normal epithelial cell sheet and (e) small lentoid bodies associated with more irregularly shaped epithelial cell sheets which accumulate significant levels of extracellular fibrillar material. (fl Large polygonal-shaped cells, (g) oriented epithelioid cells, (h) oriented fibroblast-like cells, (i) lentoid body area bordered by tibroblast-like cells, 0’) fibroblastic cells showing a loss of contact-inhibited growth present in dedifferentiated cultures of selected slow-growing N-Db, N-J, and N-Rd genotypes, and (k) irregularly shaped epithelial cells connected by extracellular filaments. (f) Pleiomorphic, swollen, and vacuolated senescent cell types present in dedifferentiated cultures of selected fast-growing HY-2 and HY-1 genotypes.

bodies still appeared in fifth passage Hy-1 cultures; later passages were not investigated. Apart from N-Rd cultures which failed to form lentoid bodies except in primary culture, the failure of fiber differentiation was progressive; the number of days required for lentoid differentiation increased and the lentoids were smaller and fewer in number with each succeeding passage (Fig. 1d). These effects were less marked in the fast-growing cultures. Thus third and fourth passage Hy-2 and Hy-1 cultures contained about 40% of the number of lentoid bodies found in primary culture, but second passage N-Db and N-J cultures had only 20% of the number of lentoids found in primary cultures. The islets of prelentoid polygonal cells were similarly reduced in successive passages (Figs.

Lens cell ageing

333

25

1 20 1

Day

of culture

Fig. 2. Cell growth in primary mass cultures of lens epithelial cultures from day-old chicks of selected slow-growing (N-Rd, N-Db, N-J) and selected fast-growing (HY-1, HY-2) genotypes. Each point represents the mean values of three determinations, the maximum variation being +lO%. Reproduced by permission of the publisher, from Ref. [ 111.

1d and 1 e). Lentoids did not degenerate or detach from the dish, but were lost at reseeding, as they were completely unable to reattach to the culture dish. The absence of lentoids in the first 10 days or more confirmed that all lentoids formed after passage were differentiated in situ and were not passively transferred. Subcultures from cultures before lentoid differentiation showed the same changes, suggesting that the process was not due to a progressive loss of cells with lentoid-forming capacity. The growth curves in Fig. 3 are based on the number of cells on Day 28 of each passage. Since each subculture was seeded at the same initial cell density, these curves make no allowance for cumulative cell death or for the possible presence of cell subpopulations with a high growth potential. This procedure may therefore diminish the extent of the true differences between the genotypes. Nevertheless there was clearly a progressive decline in cell density at the end of each successive subculture of Hy-1 and Hy-2. From the third passsage onward, there was increasing cell heterogeneity with epithelial cells, irregular cells connected by filaments (Fig. 1e), large polygonal cells (Fig. If), and mutually oriented cells (Figs. 1g and 1 h) which were epithelial-like in early passages and fibroblast-like, with a lack of contact inhibition, in later passages (Fig. lj). Oriented cells never exceeded about 30% of the cell population. Lentoids never arose from the large polygonal cells nor the oriented cells (Figs. If-1 i). The proportion of the irregular cells with filaments increased and they were the main cell type in fifth passage cultures (Fig. 1k). By the sixth passage HY-2 cells were composed only of small groups of swollen vacuolated pleiomorphic cells (Fig. 1I) with limited growth potential and a high rate of cell death. These cells were regarded as terminally senescent since a seventh passage was attempted but the cells failed to grow and then died. The onset of senescence was less marked in

334 Patek and Clayton

Passage

numbef

Fig. 3. The effect of serial culture on the cell density of LEC from day-old chicks of selected slowgrowing (N-Db, N-Rd) and selected fast-growing (HY-1, HY-2) genotypes. The cultures were subcultured every 28 days, and each time replated at the original cell density, 3~10~~ cells. Each point represents the mean values from three separate studies, the maximum variation being +lO%; 1, primary culture. Arrows indicate complete loss of lentoid body differentiation.

Hy-I which still retained a limited capacity for lentoid body formation in the fifth passage (Fig. 1e), No cellular senescence was observed in later passages of the three slowgrowing genotypes (Fig. 3). The cell density at first fell progressively and during this period the cultures showed cellular heterogeneity similar to that found in Hy1 and Hy-2. However, from the third passage of N-Rd and the fourth of N-J and N-Db, Iibroblast-like cells formed from 30 to 70% of the cell population, and in later passages only fibroblast-like cells were found (Fig. 1~). With the appearance of the fibroblast-like cells the growth rate rose steadily and confluence was achieved rapidly in later passages. These cultures behaved as if transformed. They were terminated after the fourth (N-Rd) or fifth (N-Db) passage. Cell counts were not made for N-J but the cultures showed a growth pattern and changes in the cell types similar to those of N-Db. Overall, the onset of terminal changes was inversely related to the intrinsic growth rate of the genotype. The onset of senescence occurred later in the fastest-growing Hy-1 and the onset of transformed behavior occurred earliest in the slowest-growing N-Rd. Protein Expression The series of changes in primary culture provided further confirmation of our previous observations [9, 121of the similarity between events in vitro and in uiuo. In all five genotypes both the rate of increase of crystallin expression and the rate of lentoid formation were directly related to the genetically determined intrinsic growth rate. Figures 4-6 show the numerous changes in the level of expression all of the polypeptides, which take place in subsequent serial subcultures. Only protein profiles of late cultures of the N-Db, N-J, and HY-2 genotypes are shown but the

Lens cell ageing

335

Fig. 4. SDS-PAGE showing relative accumulation of water-soluble protein (A) and fluorograms showing relative levels synthesized (S). Protein (60 ug) was analyzed on each track. Cultures from day-old chicks of the selected fast-growing HY-1 genotype were subcultured and protein profiles were analyzed at 2%day intervals. X, day 14; 1, primary culture. DO, water-soluble proteins present in lens fiber masses from day-old chicks, HY-1 genotype, 200 pg. Molecular sizes are indicated in kilodaltons, and the sizes of the crystallin polypeptides shown are 6 (48 kDa), p, (34 kDa), /I2 (26 kDa), /3, (24 kDa), fi4 (23.5 kDa), p5 (23 kDa), p6 (22 kDa), a, (20 kDa), and a2 (19 kDa). Electrophoresis was routinely terminated once the gel front (bromphenol blue dye) moved off the gel so that a2 (19 kDa) migrated to about 1 cm from the end of the gel.

sequence of change in earlier passages was similar to that of Hy-1 cultures. Crystallin expression was progressively lost during serial subculture, but the rate of loss was inversely related to the intrinsic growth rate of the genotype. The major changes in crystallin composition are summarized in Fig. 7. All the crystallins, except for the /3, (34 kDa) crystallin, were promptly lost after the first passage in N-Rd (Fig. 5) but all the crystallins, except &crystallin, persisted until the fifth passage in HY-1 cultures (Fig. 4). Only second passage cultures of the faster-growing HY-1 genotype retained a level of crystallin accumulation equaling that of primary cultures, and only these cultures showed the same progression of changes in crystallin subunit composition from 14 to 28 days as they did in primary cultures; for example a relative increase in the a20: al9 kDa ratio and a shift from a high abundance of pj (24 kDa) and p5 (23 kDa) to /3J (24 kDa) and p6 (22 kDa) (see [9, 111). The sequence of crystallin loss was, however, broadly similar in all genotypes and was generally paralleled by the changes in rates of synthesis. &Crystallin expression was lost first (Fig. 7). No 6crystallin was detected in fourth passage Hy-1 cultures, third passage Hy-2 and N-Db cultures, or in second passage N-J or N-Rd cultures. However, trace amounts of d-crystallin were still being synthesized in fifth passage Hy-1 and Hy2 cultures, and in third passage N-Db cultures. No d-crystallin synthesis was detected in sixth passage Hy-2 cultures, fourth passage N-Db cultures, or second passage N-J cultures, and no d-crystallin was detected in second passage N-Rd

336 Patek and Clayton

4

4’-

AD

Fig. 5. SDS-PAGE showing water-soluble proteins (60 ug per lane) present in LEC from day-old chicks of the selected slow-growing N-Rd genotype. Culture were subcultured and the protein profiles were analyzed at 2%day intervals: 1, primary culture; a, gel lane overloaded with 600 ug protein; AD, water-soluble proteins present in fiber masses from g-week-old adult chicks, N-Rd genotype, 200 pg. Molecular sizes are indicated in kilodaltons; see Fig. 4 for sizes of crystallin polypeptides.

cultures and was seen only faintly in fourth passage cultures after overloading of the gel lane. In all genotypes except N-Rd, the loss of d-crystallin was accompanied by a sharp fall in the levels of ,$ (26 kDa) and p3 (24 kDa) (Figs. &6). The two a-crystallin polypeptides and &, ps, and p6 (22, 23 and 24 kDa P-crystallins) were then lost together, in a subsequent passage (except in the case of N-Db when ps (23 kDa) was lost a little later than the others). The /3, (34 kDa) and p2 (26 kDa) subunits were the most persistent crystallins in all genotypes, and were the major crystallins still detectable in second, fourth, and fifth passage cultures of N-Rd, N-Db, and Hy-2, respectively (Figs. 5, 6a, and 6b). Although p1 (34 kDa) was more abundant than pZ (26 kDa) and was the most persistent in N-J and N-Rd, ,& (26 kDa) was more persistent in N-Db and both polypeptides were lost together in Hy-1 (Figs. 4, 5, 6b, 6c, and 7). Previously we have found that the loss of 8-crystallin synthesis is due to a cessation of &crystallin RNA transcription [ 191.However, even when there is no further detectable accumulation of a- and /3-crystallins, fluorography of the gels showed that most were still being synthesized, at low levels, even as late as in sixth passage senescent Hy-2 cultures (Fig. 6a), fifth passage transformed N-Db cultures (Fig. 6 b), and fourth passage transformed N-J cultures (Fig. 6 c) while heavy overloading of the gel showed GI- and P-crystallins in fourth passage transformed N-Rd cultures (Fig. 5). However while fifth passage Hy-1 cultures synthesized only trace amounts of p6 (22 kDa) it was clear that this component was not expressed in fifth passage Hy-2 cultures or third passage N-Db and N-J cultures and no p6 (22 kDa) was detectably accumulated in fourth passage N-Rd cultures (even when the gel lane was overloaded). As has been reported [193 the loss of p6 (22 kDa) expression in cultures of all genotypes was accompanied by

Lens cell ageing

DOti5

5x5 s

A

A

S

S

337

66 S

A

A

S

Fig. 6. SDS-PAGE showing relative accumulaton of water-soluble protein (A) and fluorograms to show relative levels synthesised (S). Protein (60 ug) was analyzed on each track. Cultures from dayold chicks in (A) selected fast-growing, HY-2 genotype and in (B and C) selected slow-growing N-Db and N-J genotypes, respectively. Cultures were subcultured and the protein profiles were analyzed at 2%day intervals. X, Day 14; 2, second passage; DO, water-soluble proteins present in day-old chick fiber masses, 200 ug, HY-2 genotype. Molecular sizes are indicated in kilodaltons; see Fig. 4 for sizes of crystallin polypeptides.

the expression of a 22.5K polypeptide which migrates between ps (23 kDa) and p6 (22 kDa) (Figs. 4-6). This component was detected by the third passage in N-Db, N-J, and N-Rd cultures but not until the fifth passage in Hy-1 and Hy-2 cultures: a 22.SkDa component also appears in very old cultures of chick embryo neural retina (Patek, Jeanny and Clayton, manuscript in preparation). A comparison at 14 and at 28 days shows that 34 and 26 kDa P-crystallins accumulate through the culture period of fifth passage Hy-2 cultures, although no lentoids are formed and growth is poor (Fig. 6a). The loss of crystallin expression and of lentoid-forming capacity does not appear to be related to the decline in cell

338 Patek and Clayton

Fig. 7. Crystallin polypeptide (m) present or ( absent. Summary of relative changes in crystallin polypeptide accumulation of serially subcultured LEC from selected fast-growing HY-1 and HY-2 and slow-growing N-Db, N-Rd, and N-J chick strains. Cultures were analyzed at 2%day intervals for crystallin by SDS-PAGE (60 ug protein per lane), and passaged at 28-day intervals. I, Primary culture. Some of the data were compiled from Figs. 4-6. Molecular sizes are indicated in kilodahons. Information regarding /$ (23.5 kDa) was omitted since this polypeptide was generally poorly resolved and present at only trace levels.

number. For example, when Hy-1 and N-Rd cultures are each at the point when cell number has declined to 50% of the number in 28-day primary culture, Hy-1 continues to form lentoids but N-Rd does not (Fig. 3). Furthermore, although second and fourth passage cultures of N-Rd and Hy-1 genotypes respectively contained an equal number of cells, only the latter formed lentoids and accumulated appreciable amounts of crystallin (Figs. 3-5). Finally, no lentoids developed

Lens cell ageing

A

339

B

Fig. 8. SDS-PAGE showing water-soluble proteins (60 pg per lane) present in (A) differentiated primary LEC or (E) dedifferentiated third passage LEC showing abundance of low molecular weight protein stain (
in second passage cultures of N-Rd or in third passage N-J and N-Db even when cultures were maintained for 50-70 days when the cell density was equal to that of lentoid-forming cultures of the previous passage (Fig. 3). As crystallin accumulation declines, the cellular content of the 43 kDa actin rises (Figs. 5-6). Dedifferentiated cultures of all genotypes also expressed relatively high levels of components with the following molecular weights: 21, 31, 37, 40, 46, 57, 61, 74, and 92 kDa. The 57 kDa component may be vimentin [20]. Although 60 ug protein was loaded onto each track, very little protein was seen in gel lanes from terminally dedifferentiated LEC, which were electrophoresed for the normal time and very little stain was found in the stacking gel. This suggests that the bulk of the protein in late cultures may be of low molecular weight (<18 kDa), which would be lost from the gel, and in fact large numbers of low molecular weight components were seen in gels where electrophoresis was terminated well before the 19 kDa a-crystallin was nearing the bottom of the gel (Fig. 8). a- and /3- but not 6-crystallin antigenicity was found in these cultures by hemagglutination inhibition (Zehir, unpublished data): taken together this suggests that crystallins are synthesised at low levels in late cultures and are rapidly degraded in size. DISCUSSION We found previously that the changes in crystallin expression during long-term primary cultures were the same in all these five genotypes, but occurred more 23-888332

340 Patek and Clayton rapidly in fast-growing strains [ Ill. Here we have examined genetic effects on the process of cellular ageing in vitro. During a period of up to 6 months, when cultures were maintained by serial passage, the rate of dedifferentiaton, defined by the loss of crystallin expression and of lentoid-forming capacity, is also found to be governed genetically, but it is inversely related to the intrinsic growth rate of the LEC in vitro. As in the case of differentiation, however, the sequence of changes in crystallin expression during dedifferentiation remain very similar in all genotypes. In all cases Scrystallin is the first to be lost, and this loss occurred earliest in the slow-growing strains, especially the slowest (N-Rd), and it occurred latest in the fast-growing strains, especially the fastest (Hy-1). A group of crystallins, comprising the two a-crystallins, and the 24-, 23-, and 22 kDa P-crystallins were next lost: in the same passage as &crystallin in the slowest strain, one passage later in the other two slow-growing strains, two passages after d-crystallin in the fast-growing Hy-2 strain, and they were still detectable in lifth passage cultures of the fastest growing Hy-I strain. In all genotypes /?i (34 kDa) and p2 (26 kDa) were the most persistent, and we have previously found that these crystallins are most resistant to loss following treatment with a carcinogen [9]. They became undetectable in the same passage in N-Rd and Hy-2, but /Jz(26 kDa) was the most persistent in N-Db and the PI (34 kDa) in N-J. The effect of culture ageing on crystallin expression appears to be differential, yet the tendency to maintain a similar series of changes overall in all five genotypes, from differentiation in primary culture to dedifferentiation, suggests some degree of programming in crystallin gene activation or repression, irrespective of eventual cellular senescence or transformation. The 34-kDa /3-crystallin is associated with fiber cell elongation in rho [21] yet here we find it in dedifferentiated cultures showing no fiber formation. It may be that fiber cell formation and lens transparency require the coexpression of several crystallins since several types of non-lens cells with different morphologies express appreciable levels of 6-crystallin alone [22]. While the loss of &crystallin protein is accompanied by a cessation of transcription of d-crystallin RNA [ 191,the a- and @crystallins except for &, (22 kDa) were still apparently being synthesized even though the levels of protein accumulation were too low for detection on the gels. Except in N-Rd the loss of 6crystallin occurred before the loss of lens-forming capacity, but in all strains the other crystallins were still being synthesized after lentoids were no longer formed and they could still be detected at trace levels in both senescent and transformed cultures. Previously we have reported that a-crystallin synthesis is not detectable in dedifferentiated third passage cultures 1193.However, in the present study, a longer exposure time and a higher level of labeling show that a-crystallin synthesis persists in long-term LEC. Persistent low levels of crystallins have been reported in some mammalian long-term lens cell lines: P-crystallins in rat and calf cell lines [23-251, a- and ycrystallins in calf cell lines [26], a- and/I-crystallins in bovine and rabbit LEC [17, 18, 27-301 and in virally infected rat cell lines [31], y-crystallin in mouse cell lines

Lens cell ageing

341

[32], and a-crystallin in a tumor derived from bovine LEC [30] and in rat and rabbit LEC [18, 331. The ageing changes appear to be programmed rather than random, since irrespective of growth rate or cell morphology, all five lines show similar behavior with respect to the crystallins and the same crystallins are still synthesized, albeit at a very low level, in late cultures of all five genotypes. In all cases there was a gradual loss of high molecular weight components and an increase in components of less than 19 kDa molecular weight, suggesting that degradative changes reported in mammalian lens cells [2, 34-381 can also occur in ageing chick lens cells. An increase either in actin or in vimentin, another cytoskeletal protein, has been noted in ageing mammalian lens cells in vitro [30, 39, 401and in aged rabbit lens cells in uiuo [41] and it has been proposed that high levels of actin characterize elongating cells and in cells that have ceased division [39, 42441. However, we find that actin becomes the major component both in the senescent epitheloid cells, which have virtually ceased division, derived from the two fastgrowing genotypes and in the rapidly dividing, fibroblast-like cells derived from the three slow-growing genotypes. Despite the overall similarities in the changes in polypeptide content, a striking difference between these genotypes is the eventual senescence of both the fastgrowing strains and the transformed behavior of all three slow-growing strains. These five genetically unrelated strains were obtained from geographically distinct and independent sources, and all the strains were maintained, and breeding records kept, under strictly controlled conditions. All strains have been highly inbred for long periods, three of them for over 40 years. They differ from each other in a range of other characteristics besides the growth rate of the LEC [12-161. These data suggest that the difference in terminal behavior of fast- and slow-growing genotypes may not be random, and that it merits further investigation. It seems possible that both the accelerated differentiation and the delayed dedifferentiation in Hy-1 and Hy-2 are related to the high growth rate, which is not reduced during embryonic development, as it is in normal strains [ 12, 451, so that in several respects Hy-I and Hy-2 behave as if they are younger than cells of other genotypes of corresponding age. The difference between Hy-1 and N-J cells in differentiation capacity is also manifest in clonal cultures [5]. The production of fibroblast-like cells from LEC is important in the slow-growing strains and may reflect an innate capacity of LEC. Cultures of each genotype were set up ab initio on two separate occasions and the patterns of change in cell properties and crystallin expression were reproducible. The method of isolation of the lens cells precludes mesenchymal contamination, and the synthesis of low levels of crystallins confirms their origin. The appearance of fibroblast-like cells, within the intact lens capsule or the synthesis of extracellular striated collagen, which normally characterizes mesenchyme cells rather than LEC, have been reported in cases of human lens pathology [46-49] in a mouse mutant, Sey [50], and LEC assume fibroblastic characteristics when grown in collagen matrix in vitro [51]. Fibroblast-like cells have also been

342 Patek and Clayton reported in long-term cultures of mammalian LEC and in virally transformed mammalian and chick lens cells cultures [23, 25, 31, 39, 43, 52-561. However, this morphology does not necessarily reflect the transformed state, since some transformed cultures have an epithelial morphology [17, 23, 26, 57, 581, and the epitheloid and tibroblast-like forms may be interconvertible. We have not tested the properties of the fibroblast-like cells further, and cannot therefore say whether they exhibit other characteristics of virally transformed chick LEC [531. The data presented here suggest that the accumulation of random changes is not a significant factor of cellular ageing of lens cells in vitro, and that the intrinsic rate of cell growth has an important regulatory effect on the rate of ageing. Recently, fibroblast cultures have also been found to undergo a precise series of changes in gene expression during ageing [59,60]. The chick LEC system and the genetic variants available would appear to provide a system particularly suitable for studies of growth-rate related changes during ageing, and for examination of the relationship between cell behavior and genotype. We are grateful to the Cancer Research Campaign who supported the early stages of this work, the Medical Research Council and the British Foundation for Age Research for their support, and to H & N Pfizer, Ltd., Dunbar, Scotland, Ross Poultry Products, Newbridge and Dumfries, Scotland, and the Poultry Research Centre, Roslin, Scotland, for the supply of day-old chickens. We thank Mr. F. Johnson for assistance with photography, Mrs. L. Dobbie for drawing Fig. 7 and Jackie Bogie, Anne Brown, and Jean Gardiner for typing the manuscript.

REFERENCES 1. Clayton, R. M. (1974) in The Eye (Davson, H., Ed.), pp. 399-494, Academic Press, London. 2. Harding, J. J., and Crabbe, M. J. (1984) in The Eye (Davson, H., Ed.), Vol. lB, pp. 207-440, Academic Press, New York. 3. De Pomerai, D. I., Clayton, R. M., and Pritchard, D. J. (1978) Exp. Eye Res. 27, 365. 4. De Pomerai, D. I., Pritchard, D. J., and Clayton, R. M. (1977) Deu. Biol. 60, 416. 5. Eguchi, G., Clayton, R. M., and Perry, M. M. (1975) Deu. Growth Differ. 17, 395. 6. Menko, A. S., Klukas, K. A., and Johnston, R. G. (1984) Deu. Rio/. 103, 129. 7. Okada, T. S., Eguchi, G. and Takeichi, M. (1971) Deu. Growth Differ. 13, 326. 8. Okada, T. S., Eguchi, G., and Takeichi, M. (1973) Deu. Biol. 34, 321. 9. Patek, C. E., and Clayton, R. M. (1985) Exp. Eye Res. 40, 357. 10. Patek, C. E., Vornhagen, R., Rink, H., and Clayton, R. M. (1986) Exp. Eye Res. 43, 29. 11. Patek, C. E., and Clayton, R. M. (1986) Exp. Eye Res. 43, 1111. 12. Clayton, R. M. (1979) in Mechanisms of Cell Change (Ebert, J. D., and Okada, T. S., Eds.), pp. 129-167, Wiley, New York. 13. Clayton, R. M. (1985) in The Ocular Lens: Structure, Function and Pathology (Maisel, H., Ed.), pp. 61-92, Dekker, New York. 14. Odeigah, P., Clayton, R. M., and Truman, D. E. S. (1979) Exp. Eye Res. 28, 311. 15. Odeigah, P., Clayton, R. M., and Truman, D. E. S. (1985) Curr. Eye Res. 4, 161. 16. Randall, F., Truman, D. E. S., and Clayton, R. M. (1979) Genetic Res. 34, 203. 17. Reddan, J. R., Dziedzic, D. S., Mostafapour, M. K.. McGee, S. J., and Schwartz, C. A. (1983) Curr. Eye Res. 2, 633.

18. Rink, H. (1984) Monogr. Deu. Biol. 17, 94. 19. Patek, C. E., and Clayton, R. M. (1986) Exp. Eye Res. 43, 595. 20. Ellis, M., Alousi, S., Lawnczak, S., Maisel, H., and Welsh, M. (1984) Exp. Eye Res. 38, 195. 21. Ostrer, H., Beebe, D. C., and Piatigorsky, J. (1981) Deu. Biol. 86, 403. 22. Clayton, R. M.. Jeanny, J.-C., Bower, D. J., and Errington, L. H. (1986) Curr. Top. Deu. Biol. 20, 137. 23. Hamada, Y., Watanabe, K., Aoyana, H., and Okada, T. S. (1979) Deu. Growrh Differ. 21, 205. 24. Rink, H., and Vomhagen, R. (1980) In Vitro 16, 277.

Lens cell ageing

343

25. Van Venrooij, W., Groeneveld, A., Bloemendal, H., and Benedetti, E. (1974) Exp. Eye Res. 18, 517.

26. Weinstein, B. I., Schwartz, J., Lonial, H., Dominguez, M. O., Gordon, G. G., Hochstadt, J.. Southern, D. B., Dunn, M. W., and Southern, A. L. (1982) Exp. Eye Res. 34. 71. 27. Courtois, Y., Counis, M. F., Laurent, M., Simonneau, L., and Treton, J. (1978) Interdiscip. Top. Geront. 12, 2.

28. Reddan, J. R., Friedman, T. B., Mostafapour, M. K., Bondy, R. L., Sutherland, S. H., McGee, S. J., and Goldberg, E. M. (1981) Vision Res. 212, 11. 29. Rink, H. (1978) Interdiscip. Topics, Geronf. 12, 24. 30. Simonneau, L., Herve, B., Jacquemin, E., and Courtois, Y. (1983) Exp. Cell Res. 145, 433. 31. Miller, G. C., Blair, D. G., Hunter, E., Mousa, G. Y., and Trevithick, J. R. (1979) Deu. Growfh Differ.

21, 19.

32. Russell, P., Carper, D. A., and Kinoshita, J. H. (1978) Invest. Ophrhulmol. Vis. Sci. 17, 568. 33. Reddan, J. R., Chepelinsky, A. B., Dziedzic, D. C., Piatigorsky, J., and Goldenberg, E. M. (1986) Differentiation

33, 168.

34. Alcala, J., and Maisel, H. (1985) in The Ocular Lens, Structure, Function and Pathology (Maisel, H., Ed.), pp. 169-222, Dekker, New York. 35. Dunia, I., Lien, D. N., Manenti, S.. and Benedetti, E. L. (1985) Curr. Eq’e Res. 4, 1219. 36. McFall-Ngai, M., Horwitz, J., Ding, L.-L., and Lacey, L. (1986) Curr. Eye Res. 5, 387. 37. Ringens, P. J., Hoenders, H. J., and Bloemendal, H. (1982) Exp. Eye Res. 34, 201. 38. Thomson, J. A., and Augusteyn, R. C. (1985) Exp. Eye Res. 40, 393. 39. Ramaekers, F. C. S., Hukkelhoven, M. W., Groeneveld, A., and Bloemendal, H. (1984) Biochim. Biophys. Acra 799, 221.

40. Ringens, P., Mungyer, G., Jap, P., Ramaekers, F. C. S., Hoenders, H., and Bloemendal, H. (1982) Exp. Eye Res. 35, 313. 41. Bagchi, M., Caporale, M. J., Wechter, R. S., and Maisel, H. (1985) Exp. Eye Res. 40, 385. 42. Macieira-Coelho, A., Liepkalns, C. I., and Pavian-Duttilleul, F. (1986) in Modern Trends in Ageing Research (Courtois, Y., Faucheux, B., Forette, B., Knook, D. L., and Treton, J. A. Eds.), Colloque, INSERM: EURAGE 147, pp. 41-54, J. Libbey Eurotext, Paris. 43. Ramaekers, F. C. S., Jap, P., Mungyer, G., and Bloemendal, H. (1983) Curr. Eye Res. 2, 169. 44. Wang, E., and Gunderstein, D. (1984) Exp. Cell Res. 154, 191. 45. McDevitt, D. S., and Clayton, R. M. (1979) J. Embryol. Exp. Morphol. 50, 31. 46. Font, R. L., and Brownstein, S. (1974) Amer. J. Ophthalmol. 78. 972. 47. Hayes, B. P., and Fisher, R. F. (1981) Curr. Eye Res. 1, 85. 48. Henkind, P., and Prose, P. (1967) Amer. J. Ophthalmol. 63, 768. 49. Pau, H., and Caesar, R. (1967) Graefe’s Arch. C/in. Exp. Ophthalmol. 171, 327. 50. Clayton, R. M. (1970) in Topics in Developmental Biology (Moscona, A., and Monroy, A., Eds.), Vol. 5, pp. 115-180, Academic Press, London. 51. Greenburg, G., and Hay, E. D. (1986) Dev. Eiol. 115, 363. 52. Creighton. M. O., Mousa, G. Y., and Trevithick, J. R. (1976) Differenfiution 6, 155. 53. Jones, R. E., De Feo, D., and Piatigorsky, J. (1981) Vision Res. 21, 5. 54. Mungyer, G., and Jap, P. H. K. (1978) Interdiscip. Top. Gerontol. 12, 13. 55. Reddan, J. R., McGee, J. J., Goldenberg, E. M.. and Dziedzic, D. C. (1983) Cur-r. Eye Res. 2, 393. 56. Rink, H., Vornhagen, R., and Koch, H. (1980) In Vitro 16, 15.

57. Courtois, Y., Simonneau, L., Tassin, J.. Laurent. M., and Malaise, E. (1978) Differentiarion

10,

23. 58. Arruti. C., and Courtois, Y. (1978) Exp. Cell Res. 117, 283. 59. Smith, J. R., and Pereira-Smith, 0. M. (1986) in Modern Trends in Ageing Research (Courtois,

Y., Faucheux, B., Forette, B., Knook, D. L., and Treton, J. A., Eds.). Colloque INSERM: EURAGE 147, pp. 55-63, J. Libbey Eurotext, Paris. 60. Stein, G. H., and Atkins, L. (1986) Proc. Nat/. Acad. Sci. USA 83, 9030. Received February 8, 1987 Revised version received July 14. 1987

Prmted

in Sweden