Exp.

Eye RPS.

(1990) 50, 345-354

Age-related Long-term

Changes in the Response of Chick Lens Cells Culture to Insulin, Cyclic AMP, Retinoic Acid Bovine Retinal Extract C. E. PATEK*

Department

of Genetics,

(Received

AND

R. M. CLAYTONi

University of Edinburgh, The Kings Edinburgh EH9 3JN. U.K.

23 February

During and a

1989 and accepted

in revised

Buildings,

West Mains Road,

form 8 May 1989)

We have reported that l-day-old post-hatch chick lens epithelial cells lose the capacity for lentoid body formation and &crystaILin expression during long-term serial subculture, although they continue to synthesize, but not to accumulate, a- and Bcrystallins, even in cells with a transformed phenotype. Here we present evidence that dedifferentiation may reflect an age-related change in the capacity for response to regulatory signals. We have tested the capacity of these cells in serial subcultures to respond to agencies which affect lens cell growth and differentiation in primary culture: retinoic acid (RA), insulin, CAMP and bovine retinal extract (BRE). Secondary cultures responded only to RA and BRE, by an increase in lentoid formation and by cz- and p-accumulation, while RA also restored S-crystalhn expression. Later cultures showed no such responses. The results suggest that the process of lens ceil dedifferentiation may, at first, be reversible but later becomes irreversible, despite the continuing persistence of low levels of crystallin expression. Key words: ageing : chick lens epithelium : crystallin : dedifferentiation ; SDS-PAGE ; retinoic acid : insulin ; CAMP ; retinal extract : S-crystallin mRNA 1. Introduction In the vertebrate lens the rate of terminal lens fibre cell differentiation and crystallin expression in successively formed fibres are both modified by age. The study of these age-related changes is important since the proliferation and differentiation of these cells regulate lens function and they are modified in certain pathologies (see Harding and Crabbe, 1984; Rink, 1984; Piatigorsky, 1987; Patek and Clayton, 1988). Age-related changes in the available external signals, or in the capacity for response to these signals, are possible mechanisms. It is tests of the second possibility that we report here. Lens epithelial cells in both birds and mammals can differentiate in vitro to form crystallin-rich aggregates of lens fibre cells termed lentoid bodies, but the study of ageing in human and other mammalian lens cells is restricted by their poor capacities for crystallin expression and lentoid formation when cultured free of the lens capsule (see Jacob, 198 7 ; Lipman and Taylor, 198 7 ; Muggleton-Harris and Higbee, 198 7 ; Arita, Lin and Reddy, 1988: Reddy et al, 1988), although some cell lines still continue to synthesize crystallins (Reddan et al., 1986; Nagineni and Bhat, 1989). Findings based on established lens cell lines may reflect adaptation to the culture conditions rather than in vivo changes (see Lens&a et al., 1982). Chick lens epithelial cell cultures (LEC) may offer a more suitable * Current address: Department of Pathology, University Edinburgh.The MedicalSchool. Teviot Place, Edinburgh EH8 [J.K. t .For correspondence. 00144835/90/040345+

10 $03~00/0

of 9AG.

in vitro model for studying cellular ageing : the morphology of fibre differentiation is similar in vivo, in vitro (see Menko et al, 19 8 7) and in primary culture : the sequence of changes in the expression of the d-, /3and S-crystallin subunits, the non-crystallins proteins and the membrane proteins are all similar to those occurring in vivo (Patek and Clayton, 1985. 1986a; Patek et al., 1986) and lentoids formed at different stages in serial subcultures show the same age-related selective loss of S-crystallin gene expression as do cortical fibres in vivo (Patek and Clayton, 1986b). Repeated subculture leads to eventual loss of capacity to form lentoids and although a- and /j’-crystallins are still synthesized, they are not accumulated ; high molecular weight proteins are lost and are replaced by proteins of low molecular weight, and the cells express high levels of actin. Since many of these changes also typify the older adult lens (see Harding and Crabbe, 1984; Alcala et al., 1988 ; Srivastava and Srivastava, 1989) we have argued that dedifferentiation may reflect the age-related change in the capacity for differentiation found in vivo (Patek and Clayton, 1986b, 1988). Although lens development is regulated by proximity to the retina in vivo the precise signal or combination of signals involved is unknown. The growth and development of lens cells is affected by a variety of hormones and growth factors (some isolated from ocular tissues) including insulin, insulin-like growth factors, platelet-derived growth factor, lentropin, lenmofin. fibre differentiation factor and fibroblast growth factors (FGFs) (see Beebe, 1985; Beebe et al., 19 8 7 ; Chamberlain and McAvoy, 198 7 : Brewitt and 0 1990 Academic Press Limited

C. E. PATEK

Clark, 1988; Mascarelli, Courtois and Arruti, 1989). The age-related changes in lens development imply either a change in the output of regulatory factors, or in the response of lens cells to such factors. There are no reports to date of the response of lens celIs to regulatory factors during in vitro ageing, and the relationship to the differentiation potential of the cells. We have therefore tested the response to lens cells at different stages of subculture to retinoic acid (RA) and insulin, both of which are potent stimulators of chick lens cells differentiation in primary cultures (Piatigorsky, 198 1; Patek and Clayton, 1986c, unpubl. res.), and to CAMP and to bovine retinal extract (BRE, which contains FGF activity), both of which affect the growth, shape and differentiation of mammalian LEC (see Creighton and Trevithick, 19 74 ; Chamberlain and McAvoy, 198 7 ; Mascarelli et al., 1989) We have found that older cultures, as they dedifferentiate, lose the capacity to respond to these agencies. We have also found that the rate of dedifferentiation is not related to the degree of prior differentiation of these cells and that the loss of Scrystallin expression is reversible during the initial stages of dedifferentiation.

2. Materials

AND

R. M. CLAYTON

cultures were analyzed on day 28. This somewhat high dose level of insulin was found to be effective in promoting lentoid formation and crystallin expression in primary cultures (Patek and Clayton, unpubl. res.i All agents were obtained from the Sigma Chemical Company Ltd, Poole, U.K., except for BRE which was a generous gift from Dr Yves Courtois, INSERM, Paris (Arruti and Courtois, 1978). CAMP and BRE were added directly to the culture medium, but RA in 9 5 “/o ethanol and insulin in 0.01 N HCl were added at a final solvent concentration of 0.1% (v/v). Insulin- and RA-treated cultures were also supplemented with 95 % ethanol and 0.01 N HCl (0.1 % v/v). respectively, and control, CAMP- and BRE-treated LEC all contained 0.1% (v/v) of each of the solvents alone. These solvent concentrations were found to have no effect on the growth, morphology or the protein composition of the cultures. Cultures were labelled with [3H]mixed amino acids (50 ,uCi ml-’ 3 hr-‘) Protein and RNA Analysis

Water-soluble proteins were analyzed by SDS-polyacrylamide gel electrophoresis (SDS-PAGE) and fluorography (Patek and Clayton, 1985). Total cellular RNA was analyzed by Northern transfer using a cloned Scrystallin cDNA, pM56 (Patek and Clayton, 1986b).

and Methods

Cell Culture

We chose to examine lens epithelial cell cultures (LEC) from the slow-growing N-Rd strain, since the sequence of changes in crystallin expression during in vitro ageing is similar to those of other strains examined but it occurs more rapidly (Patek and Clayton, 1988). Lens epithelia were dissociated and cultured as described previously (Patek and Clayton, 1985) using one single batch of foetal calf serum. The effect of prior differentiation of the cultures on the rate of lens cell dedifferentiation was assessed by passaging either from ‘/-day undifferentiated primary cultures (LEC- 7) or from 28-day cultures containing numerous lentoids (LEC-28). LEC-7 and LEC-28 were then subcultured at 7- and 28-day intervals, respectively, and, in each case, the cultures were replated at the original cell density. Most lentoids were removed from the dissociated cells by sedimentation before replating, but any remaining lentoids do not attach and were lost at the first medium change. Sister cultures from those passaged at 7 days were maintained for 28 days to assess their differentiation capacity. A parallel set of cultures of second, third and fourth passage LEC-28 were treated with cyclic adenosine 3’, 5’-monophosphate (10 ,ug ml-l, CAMP), /3-all-truns retinoic acid (1 ,ug ml-‘, RA), a bovine retinal extract (50 ,ug ml-l BRE) or insulin (6 ,ug ml-‘, bovine pancreas crystalline), every 3 days between days 3 and 2 7 inclusive, and the proteins from treated and control

3. Results (a) Dedifferentiation of Lens Cells from Previously Diferentiated and Undiflerentiated Cultures: Evidence for the Non-random Character of the Changes in Crystallin Expression

The series of changes occurring during differentiation in primary cultures and during dedifferentiation in serial subcultures has been described in detail (Patek and Clayton, 1985, 1986b, 1988) and is confirmed in these experiments. LEC-7 and LEC-28 subcultures showed a similar sequence of change in morphology and protein expression, but cells from LEC-28 dedifferentiated more rapidly. Primary cultures were initially composed of flattened epithelial cells, [Fig. l(A)]. Islets of pre-lentoid polygonal-shaped cells appeared by day 9, and lentoids with characteristic bottle cells formed by day 16 [Fig. 1 (A), (B)], but about 60 % of the cell sheet still remained undifferentiated by day 28. Second-passage LEC-28 contained about 50% fewer cells than primary LEC and were composed of epithelial cells connected by extracellular fibrillar material [Fig. 1 (E)] and small lentoids [Fig. 1 (D)] which were more numerous in LEC-7. Third-passage LEC-28 contained about 70% fewer cells than primary LEC and were composed of pleiomorphic epithelioid cells, large polygonal-shaped cells and oriented cell types [Fig. 1 (F), (G)]. Small lentoids were still detected in third-passage LEC-7. Fourth-passage LEC-28 contained about three times more cells than primary

CHICK

LENS

CELL

347

AGEING

FIG. 1. Phase contrast photomicrographs ( x 160) showing chick LEC passaged at 28-day intervals and the effect of RA treatment (1 /Lg ml-l) every 3 days between days 3 and 2 7 inclusive. Control cultures : (A) 7-day undifferentiated primary LEC showing flattened epithelial cells; (B) 14-day primary LEC showing ‘islets’ of small polygonal-shaped cells bordered by developing lentoid bodies: (C) 28-day primary LEC showing mature lentoid bodies with characteristic bottle cells: (D) 28-day second-passage LEC showing small lentoid bodies ; (E) 28-day third-passage LEC showing irregularly-shaped epithelioid cells and extracellular material: (F) third-passage LEC showing large polygonal-shaped cells: (G) 28-day third-passage IBC showing oriented elongated cells; (H) 28-day fourth-passage LEC showing rapidly growing fibroblast-like cells. RA-treated cultures (day 28): (I) second-passage LEC showing extensive areas of lentoid bodies : (J) second-passage LEC showing dense areas of epithelioid cells; (K) third-passage LEC showing irregularly-shaped epithelioid cells and extracellular material: (L) fourth-passage IX showing irregularly-shaped epithelioid cells with abundant extracellular material. RA-treated fourth passage LIT contained none of the fibroblast-like cells seen in control cultures. cultures and were composed of a homogenous population of rapidly growing fibroblast-llle cells [Fig. 1 (H)] but fourth-passage LEC-7 also contained some

pleiomorphic epithelial cells, and rare structures resembling small lentoids [Fig. 1 (E)]. By the fifth passage, LEC-7, like fourth-passage LEC-28, were composed entirely of Ebroblast-like cells which did not show contact-inhibited growth but still expressed crystallins. This phenotype has been reported in virally and spontaneously transformed avian and mammalian lens cells (see Hamada et al., 1979 ; Jones, DeFeo and Piatigorsky. 198 1; Ramaekers et al., 1984 ; Patek and Clayton, 1988). Subculture of both LEC-7 and LEC-28 was marked by a loss of accumulation of S-crystallin followed by a-

and finally /3-crystallin: and in both cases actin (43 kDa) was the major product present in later cultures, along with lower levels of the 46 and 58 kDa polypeptides and several in the 3540 kDa size range (Fig. 2). Cultures replated at 7 days always contained more crystallin overall than those replated at 28 days, and, unlike with LEC-28, traces of Kcrystallin were still found in second-passage LEC- 7 and a-crystallin in third-passage LBC-7. The series of changes, including the preferential loss of S-crystallin [see also Figs 3 (lane C), 4 (lane B), 6 (lane B), 7 (lane G), 8 (lane A)] and its mRNA [Fig. 9 (lanes A and D)] in aged LEC-28 confirms our previous Endings (Patek and Clayton, 1986b. 1988). We also confirm that fourth-passage LEC-28 retain a persistent

capacity to synthesize

low

348

C E. PATEK

AND

R. M. CLAYTON

levels of a- and /I-crystallins [Fig. 7 (lane G)]. Although 60 ,ug of protein was loaded on to each track, only five polypeptides in the 38-58 kDa size range could be detected in fifth passageLEC-7. Protein overloading shows that high molecular weight polypeptides are still present, including a- and /I-crystallins [Fig. 10 (lanes A-E)] as identified by the position in the gel and by haemagglutination inhibition using monospecific antisera (Zehir, unpubl. res.). The apparent loss of soluble protein in aged LEC-7 is probably due to degradation rather than insolubilization, since remains at the origins and there are no protein appreciable levels of low molecular weight ( < 19 kDa) polypeptides not found in primary cultures [Fig. 10 (lanes A, F and G)]. These changes in aged LEC-7, including the appearance of low molecular weight polypeptides are similar to these found previously in LEC-28 (Patek and Clayton, 1988). We may therefore conclude that they are not random, since they occur each time cultures are established from l-day-old chick lens.

ABCDEFGHI FIG. 2. SDS-PAGE showing water-soluble proteins (60 ,ug) present in : (A) 28-day second-passage LEC-28 ; (B) 7-Day primary LEC; (C-D) 28-day third- and fourth-passage LEC-28, respectively: (E) l-day-old chick LFM: (F) 5 ,ug chick muscle G-actin, 43 kDa (G-I) 28-day third-, secondand fourth-passage LEC-7. respectively, Molecular sizes are indicated in kDa.

ABCDE

(b) The Effects of RA, Insulin, CAMP and BRE on the Processof Dediflerentiation LEC-28 cultures were treated with RA, BRE, CAMP and insulin during the second, third or fourth passages. Continuous treatment from the primary

ABC

D

ABC

FIG. 3. SDS-PAGE showing water-soluble proteins (60 pug) present in : (A) l-day-old chick LFM: (B) 5 ,ug chick muscle G-actin; (GE) 28-day second-passage LEC-28 treated with BRE (50 pg ml-l) or insulin (6 pg ml-‘) every 3 days between days 3 and 2 7, inclusive during the second passage: (C) control: (D) BRE-treated; (E) insulin-treated. Molecular sizes are indicated in kDa. FIG. 4. SDS-PAGE showing water-soluble proteins (60 fig) present in 28-day second-passage LEC-28 treated with RA (1 pg ml-l) or CAMP (10 yg ml-l) every 3 days between days 3 and 2 7, inclusive during the secondpassage ; (A) l-day-old chick LFM; (B) control: (C) RA-treated; (D) CAMP-treated. Molecular sizes are indicated in kDa. FIG. 5. SDS-PAGE showing water-soluble proteins (60 rug) present in 28-day third-passage chick IX-7 treated with RA (1 pg ml-l) or CAMP (10 ,ug ml-‘) every 3 days between days 3 and 2 7, inclusive during the third passage. (A) control : (B) RAtreated: (C) CAMP-treated. Molecular sizes are indicated in kDa.

CHICK

LENS

CELL

349

AGEING

6

ac ABCDEF

ABCDEF

G

FIG. 6. SDS-PAGEshowingwater-solubleproteins(60 pug)presentin (A) LX-week adult chick LFM; (B-F) 28-day third-passage

LEC-28treated with insulin (6 ,ugml-l), CAMP (10 pg ml-‘), BRE(50 pg ml-l) or RA (1 pg ml-‘) every 3 days betweendays 3 and 2 7. inclusiveduring the third passage;(B) control ; (C)insulin-treated;(D)CAMP-treated:(E)RA-treated: (F) BRE-treated. Molecular sizesare indicatedin kDa. FIG. 7. SDS-PAGEshowingwater-solubleproteins(60 pg) presentin (A-C and F) 2%day fourth-passage LEC-28treatedwith BRE (50 ,ugml-‘), insulin (6 yg ml-l) or RA (1 yg ml-l) every 3 days betweendays 3 and 27, inclusive during the fourth passage.(A) control ; (B) BRE-treated; (C)insulin-treated; (D) l-day-old chick LFM; (E) 5 pg chick muscleG-actin, 43 kDa. (F) RA-treated. Molecular sizes are indicated in kDa. (G) fluorogram (SDS-PAGE)showing water-soluble proteins (60 pg) synthesisedfrom 3H-labelledmixed amino acids(50 $i ml-’ 3hr-‘) by 35-day fourth-passageLEC-28.

culture was avoided to guard against possible effects due either to cell selection or to desensitisation as can occur following prolonged treatment with insulin, CAMP and epidermal growth factor (Skreb et al., 1984 ; Arruti, Cirillo and Courtois, 1985 ; Chiocca, Davies and Stein, 1988). The RA and insulin treatment regimes used strongly promoted growth, differentiation and crystallin expression in the primary cultures (Patek and Clayton, unpubl. res.).

this polypeptide is S-crystallin, partly because of its size and partly because it is invariably present only in cultures containing fully processed(2 kb) 6-crystallin RNA. The absence of 6-crystallin from all other cultures even when maintained for > 28 days (by which time the cell density equalled that of dIfferentiated primary cultures) precludes cell density as the regulatory factor. RA promotes S-crystallin accumulation preferentially in primary LEC (Patek and Clayton, 1986c, unpubl. res.) but /3-crystallin still

Second-Passage Cultures (LEC-28). CAMP had no effect on growth or lentoid formation and failed to promote crystallin accumulation but caused loss of a1 (aB, 20 kDa) and also affected non-crystallins: for example treated cultures contained less actin (43 kDa) and less of the 46 kDa polypeptide than controls [Fig. 4 (lane D)]. Insulin, BRE and RA increased cell number by 30, 30 and 400%, respectively, but only RA and BRE increased the number and size of the lentoids differentiated [Fig. 1 (I)]. RA-treated LEC contained appreciably more lentoids than BRE-treated LEC and there were areas of densely packed small cells, not seen in controls [Fig. 1 (J)]. All three agencies affected non-crystallins and promoted /3-crystallin accumulation, but while RA and BRE showed preferential stimulation of expression of p3 (24 kDa), p5 (2 3 kDa) and /I’~ (22 kDa). insulin promoted the expression of p, (34 kDa) only (Figs 3 and 4). RA and BRE promoted s-crystallin, but BRE preferentially promoted a1 (aB), and only RA-treated cultures contained a 48-kDA polypeptide and detectabIe amounts of S-mRNA [Fig. 9 (lane B)]. We believe that

remains the most abundant in RA-treated second-

passage LEC. Third-Passage Cultures. (a) LEC-28. None of the agencies tested promoted fibre differentiation or crystallin accumulation in third-passage LEC-28 (Fig 6). RA also failed to affect a- and /3-crystallin synthesis or to restore S-crystallin expression [Figs 8 and 9 (lane E)] even at high concentrations (3 lug ml-‘, data not shown). Only RA promoted mitosis, the cultures achieving a cell density similar to those of differen-

tiated 28-day primary cultures. RA alone affected cell morphology: these cultures were composed of epithelioid cell types [Fig. 1 (K)] and contained none of the polygonal-shaped and ‘ elongated ’ cells found in controls [Fig. 1 (F), (G)]. However all agents still affected non-crystallins. the pattern of response being specific for each agency (Fig. 6). (b) LEC-7. Third-passage LEC-7 still formed small lentoids and had detectable levels of a- and /3crystallins. The response to CAMP and RA was

350

C. E. PATEK

R. M. CLAYTON

growing at the previous lower rate [Figs 1 (I,) and 7 (lane F)].

9

8

AND

(c) The Ejjfect of Insulin Pretreatment on Lens Cell Dedifjerentiation (LEC-7 and -35)

In a pilot experiment we examined the effect of early treatment with insulin (6 ,ug ml-‘) on the process of dedifferentiation of cells from undifferentiated and well-differentiated primary cultures (LEC- 7 and LEC3 5, passaged at 7 and 35 days, respectively). Cells which had been treated with insulin in primary culture were able to respond to insulin in second- and third-passage cultures by an increase in fibre differentiation and in the expression of a-, b- and Scrystallins (Figs 11 and 12). Thus, continuous exposure to insulin 1

A

6

ABCDE

FIG. 8. SDS-PAGE showing water-soluble proteins (60 pug) synthesized from 3H-labelled mixed amino acids (50 &i ml-’ 3hr-‘) by 28-day third-passage LEC-28 treated with RA (1 ,ug ml-‘) every 3 days between days 3 and 2 7, inclusive of the third passage. (A) control: (B) RA-treated. Molecular sizes are indicated in kDa. FIG. 9. Autoradiographs of RNA (Northern) transfers hybridized to Y2P-labelled S-crystallin cDNA clone, pM56, showing total cellular RNA (15 pg) from 28-day secondand third-passage LEC-28 treated with RA (1 ,ug ml-l) every 3 days between days 3 and 27, inclusive of the second and third passages, respectively. (A) control primary LX; (B) RA-treated second-passage LJX; (C) polysomal RNA (1 pug) from l-day-old chick LFM showing mature (2 kb) S-crystallin mRNA: (D) control second-passage LEC; (E) RA-treated third-passage LEC. DNA and RNA size markers not shown. No hybridization was evident when duplicate filters were hybridised to 32P-labelled pBR322, the plasmid vector into which cDNA was inserted to give pM56. Size of RNA is indicated in kilobases.

examined. In a pilot experiment only RA promoted lentoid formation and /3-crystallin accumulation (no data were obtained for ol-crystallin) and restored Scrystallin accumulation, but P-crystallin remained most abundant (Fig. 5). However, CAMP still affected non-crystallins : for example, the cultures contained

lower levels of actin (43 kDa) and of a 46 kDa polypeptide than controls. Fourth-Passage Cultures (LEG28). We did not test CAMP since it had failed to promote differentiation in earlier passages. Fourth-passage LEC-28 still synthesized low levels of a- and P-crystallins (Fig. 7 (lane G)) but neither insulin nor BRE had any effect on the growth, morphology or protein composition of the cultures which were now growing rapidly, and had become fibroblastic in appearance [Fig. 7 (lanes B, C)]. RA-treated cultures showed no effect on the protein composition but they remained epithelioid, and were

appears

to delay the process

of lens ceil

dedifferentiation. In the case of LEC-35, insulin treatment restricted to primary cultures only, still slowed the rate of dedifferentiation, but the secondpassage cultures contained less crystallin overall than cuitures in which

treatment had been continued in the

second passage. 4. Discussion We have confirmed our previous findings (Patek and Clayton, 1986b, 1988) that dedifferentiation of chick lens cells in serial subcultures is not due to a cessation of crystallin synthesis since fourth-passage LEC still express low levels of a- and /3-crystallins nor to changes in cell density, nor to a passive loss of earlier formed lentoids. Here we show that dedifferentiation is not due to a loss of cells with high differentiation potential due to prior differentiation of the cultures, since LEC-7 subcultures also dedifferentiate, although less rapidly than LEC-28 subcultures, possibly because LEC-7 have undergone fewer mitoses than LEC-28. For any passage, LEC-28 cultures have been dividing for a period roughly four times longer than LEC-7 of the equivalent passage number. We find that the sequence of changes in crystallin expression is similar each time serial subcultures of l-day-old chick lens cells are established de novo, whether from different genotypes (Patek and Clayton, 1988) or subcultured under different regimes (this report), and since there are also similarities between ageing chick and mammalian lens cells (Patek and Clayton,

1988) we suggest that

lens cell in general may undergo a sequence of changes many of which are non-random. This supports our earlier hypothesis that there is an intrinsic

age-related

sequence

of changes

during

differentiation and subsequent dedifferentiation (Patek and Clayton, 1985, 1986a, b, 1988). The effects of insulin and neural retina-conditioned medium on lens cell development in vitro are modified 1981: Richardson and by donor age (Piatigorsky,

McAvoy, 1988). Here we have examined the response of lens cells to regulatory factors during in vitro

CHICK

LENS

CELL

AGEING

351

A

C

B

D

F

E

G

FIG. 10. SDS-PAGE showing water-soluble proteins (60 pg except where stated) present in 2%day LEC-7; (A) fifth-passage LEC ; (B) 1 -day-old chick LFM : (C and D) as lane A but 230 and 600 pg protein, respectively: (E and F) 20- and 40-day primary LEC, respectively: (G) 40-day third-passage LEC (subcultured at 4-day intervals): OS and Or, origins of the stacking and resolving gels: Fr. gel front marked by bromophenol blue dye; LMW, low molecular weight protein (< 19 kDa). Lanes A-E are from the same gel. In lanes F and G, electrophoresis was terminated early when the gel front was 1 cm from the end of the gel. Molecular sixes are indicated in kDa.

II

A

BCDE

A

BCD

FIG. 11. SDS-PAGE showing water-soluble proteins (90 pg) present in 35day second-passage LEC (subcultured at 35day intervals) treated with insulin (6 pug ml-l) every 3 days between days 3 and 33 inclusive. (A) control LEC; (B) insulin-treated during the primary culture only: (C) insulin-treated during the second passage only: (D) insulin-treated during primary and second passages ; due to interfering stain the position of S-crystallin is not marked ; (E) lower band is 5 pug chick muscle G-actin. 43 kDa. Molecular sixes are indicated in kDa. FIG. 12. SDS-PAGE showing water-soluble proteins (90 pg) present in 3 5-day third-passage IX (subcultured at 7 day intervals) treated with insulin (6 ,ug ml-‘). (A) control LEC; (B) insulin-treated between days 3 and 33, inclusive of the third passage; (C) insulin-treated on days 3 and 6 of the primary and second passages, and every 3 days between days 3 and 33 inclusive of the third passage: (D) 5 pug chick muscle G-actin. 43 kDa. Molecular sizes are indicated in kDa.

352

ageing, and the relationship of these responses to the differentiation potential of the cells. During the process of dedifferentiation, lens cells derived from primary cultures lose the capacity for response to all the four agencies tested, but both the nature and specificity of the response and the stage at which the response is lost is specific to each of these agencies. For example, RA and BRE both promote cell growth, fibre differentiation and the accumulation of a- and /3-crystallins in secondary cultures, but RA also promotes &crystallin expression. All agencies except CAMP promoted growth of second-passage LEC, but only RA promoted the growth and affected the morphology of thirdpassage LEC. All four agencies affected the levels of non-crystallin polypeptides in the third passage but none had such effects on fourth passage cells. RA inhibited growth of fourth-passage LEC but maintained epithelioid morphology and inhibited fibroblastic cell formation. The failure of these agents to affect differentiation of later subcultures could not be due to loss of crystallin expression, since CI- and P-crystallins are still synthesized in fourth-passage cultures: nor can it be due to cellular transformation, since BRE and RA failed to influence differentiation of third-passage LEC which contain no fibroblast-like cells. Finally, it could not be due to a total failure of cellular responses, since all agencies still affect non-crystallin levels in third-passage LEC, and RA still affects growth and morphology of third- and fourth-passage LEC. The data support the view that changes in development and crystallin expression in vitro and in the ageing lens in vivo may involve changes in the spectrum of responses to signals and changes in receptors, rather than a loss of extrinsic regulatory factors. Insulin promotes growth rather than fibre differentiation in lens cells from older chick embryos (Piatigorsky, 1981), and here we noted a similar change in response to insulin between first- and second-passage LEC. We found that LEC begin to dedifferentiate before they lost the capacity for response to RA and BRE. Since insulin delayed lens cell dedifferentiation, provided that it was present from the beginning, this dedifferentiation could be due to a deficiency of fibrogenic factors in the culture medium. The various age-related changes in cellular response we report here may involve changes in receptor levels, since insulin and retinoid receptor levels in ocular tissues do change during development in vivo (Wiggert, Masterson and Coulombre, 1983: Bassas et al., 1987). Since RA promoted lentoid formation and crystallin accumulation and restored b-crystallin expression in second-passage LEC while continuous insulin treatment preserved the potential for response to insulin, we conclude that the capacity for differentiation and for S- crystallin expression is suppressed but not lost during the early stages of dedifferentiation. Reversible changes in LEC differentiation and crystallin expression have also been reported following cold shock (Creighton et al., 1981), mitotic arrest due to serum

C. E. PATEK

AND

R. M. CLAYTON

starvation (see Rink, 1984), tumour formation in ‘nude’ mice (Simonneau et al.. 1983a), cell transformation (Rink and Vornhagen, 1980; Weinstein et al., 1982) and in RSV-infected chick LEC and quail neural retina cultures when maintained at nonpermissive temperatures for viral growth (Simonneau et al, 1986; Menko and Boettiger, 1988). An important factor in reversibility may be the degree of ageing change, since we found that the effects of RA and BRE on fibre differentiation and crystallin expression were reversible at early stagesof dedifferentiation but later became irreversible, and the capacity for restoration of y-crystallin synthesis following cold shock or mitotic arrest similarly falls with in vitro age (Creighton et al., 1981; Rink, 1984). Crystallin synthesis was no longer restored in our older cultures, and other investigators have also failed to restore crystallin expression in aged cells by FGF (Simonneau et al., 1983b), glutathione, ascorbic acid, foetal calf serum and dibutyryl CAMP (van Venrooij et al.. 1974). The maintenance and promotion of Scrystallin expression by RA in our study is probably at the transcriptional level : RA specifically activates Scrystallin expression in transfected murine teratocarcinoma cells (Goto et al., 1988). and selectively affects gene transcription in other systems (Sherman, 1986; Chiocca et al., 1988: La Rosa and Gudas. 1988 ; Maviiio et al., 1988 ; Thiele, Deutsch and Israel, 1988; Bedo, Santisteban and Aranda, 1989). This effect on S-crystallin is a specific responseto RA rather than a general response of the cells to growth and differentiation. Cell growth, morphology and noncrystallin expression are still responsive to RA after the capacity for response of bcrystallin to RA has been permanently lost. The difference in cellular responseto CAMP and RA confirms that the effect of RA is not mediated by CAMP (seeMaddox and Haddox, 1988). We also confirm that BRE and RA affect lens cells differently (Baritault. Arruti and Courtois, 198 1). Lens development is affected by the retina but retina-derived factors such as FCF have only recently been identified, and several growth factors which affect lens cells are present in the aqueous and vitreous humors (Beebeet al., 1987; Brewitt and Clark, 1988; Richardson and McAvoy, 1988 ; Mascarelli et al., 1989). The retina is rich in retinoids (see Sherman, 1986) but to date there is no evidence for any free retinoids in the aqueous or vitreous, and this paper presents the first evidence for responsivenessof lens cells to RA. We cannot therefore say, at present. whether the failure of aged LEC to transcribe 6crystallin RNA in response to to RA has an in vivo counterpart, although d-crystallin transcription ceases in the mature lens. RA is both a morphogen (Maden et al., 1988; Thaller and Eichele, 1988) and a potent anticarcinogen (see Sherman, 1986). Whether these properties have any relationship to the absence of naturally occurring lens neoplasia (see Piatigorsky,

CHICK

LENS

CELL

353

AGEING

1987: Westphal, 1988) is at present uncertain. However, we did find that RA apparently prevents lens cell transformation, which normally occurs in aged LEC derived from this strain (Patek and Clayton, 1986b, 1988). This effect of RA was without effect on protein expression and so conilrms our previous finding that protein expression in aged LEC is independent of changes caused by transformation (Patek and Clayton, 1988). We suggest that chick LEC provides a valuable in vitro model for the study of cellular ageing and the regulation of crystallin expression. These cells lose the capacity for lentoid dedifferentiation and crystallin accumulation after being maintained in conditions which preclude differentiation for as short a period as 7 days. This finding clearly emphasizes the possible danger in using well-established lens cell lines as

a starting problems.

point

for investigations

of age-related

It would be of interest to investigate the relationship between RA and the apparent activation of 6-crystallin

gene expression in aged cultures and to determine whether this represents de novo induction of transcription or a change in the rate of transcription, or of processing or stability of the transcripts. Since lens

cells are constantly exposed to fibrogenic factors in vivo it would also be relevant to examine the effect of continuous treatment with RA from the primary culture on the ageing process, and determine whether the effect of RA is modified by genotype and donor age. Acknowledgments Weare grateful to the British Foundationfor AgeResearch who supported the work and Ross Poultry Products, Dumfries,U.K.. for the supply of l-day-old chickens.We also thank Dr YvesCourtois(INSRRM,Paris)for the gift of bovine retinal extract, and Mr Frank Johnstonfor assistancewith photography. References Alcala, J.. Katar. M.. Rudner, G. and Maisel, H. (1988). Human beta crystallins: regional and age-related changes.Curr. Eye Res. 7. 353-9. Arita, T.. Lin, L.-R. and Reddy, V. N. (1988). Differentiation of human lensepithelialcellsin tissueculture. Exp. Eye Res. 47. 905-10.

Arruti. C.. Cirillo,A. and Courtois,Y. (1985). An eye-derived growth factor regulatesepithelial cell proliferation in the cultured lens. Diflerentiation 28, 286-90. Arruti. C. and Courtois,Y. (1978). Morphological changes and growth stimulationof bovine epitheliallenscell by a retinal extract in vitro. Exp. Cell Res. 117. 283-92. Baritault, D., Arruti. C. and Courtois,Y. (1981). Is there a ubiquitousgrowth factor in the eye?Differentiation IS. 2942. Bassas, L., Zelenka,P. S.,Serrano.J. and dePablo,F. (1987). Insulin and IGFreceptorsaredevelopmentallyregulated in the chick embryo eyelens.Exp.CeJJ Res.168, 561-6. Bedo,G., Santisteban.P. and Aranda. A. (1989). Retinoic acidregulatesgrowth hormonegeneexpression.Nature 339.2314. Beebe, D. C. ( 1985). Ocular growth and differentiation

factors. In Growth and Maturation Factors, Vol. 3 (Ed. Guroff, G.). Pp. 37-77. Wiley: New York. Beebe, D. C., Silver. M. H., Belcher, K. S., von Wyk, J. J., Svaboda.M. E. and Zelenka,P. S. (1987). Lentropin, a protein that control lens fiber formation, is related functionally and immunologically to the insulin-like growth factors. Proc. Natl. Acad. Sci. U.S.A. 84, 2327-30. Brewitt, B. and Clark,J. I. (1988). Growth and transparency in the lens,on epithelialtissue,stimulatedby pulsesof PGDF.Science 242, 777-9. Chamberlain,C. G. and McAvoy, J. W. (1987). Evidence that fibroblast growth factor promotes lens tibre differentiation. Curr. Eye Res. 6. 1165-8. Chiocca, E. A.. Davies, P. J. A. and Stein, J. P. (1988). The molecularbasisof retinoic acidaction. 1. Binl. Chem. 263. 11584-9. Creighton.M. O., Mousa,G. Y., Miller, G. G., Blair, D. G. and Trevithick. J, R. (1981). Differentiation of rat lens epithelialcellsin tissueculture. IV. Somecharacteristics of the process,including possiblein vitro modelsfor pathogenic processesin cataractogenesis.Vision Res. 21, 25-35. Creighton,M. 0. and Trevithick, J. R. ( 1974). Effectof cyclic AMP, caffeine and theophylline on differentiation of lensepithelial cells.Nature249, 767-8. Goto. K., Hayashi. S., Shirayoshi, Y., Takeichi. M. and Kondoh, H. (1988). Exogenousbcrystallin gene expressionasprobefor differentiation of teratocarcinoma stemcells.CellDij‘er. 24, 13948. Hamada.Y., Watanabe, K.. Aoyama. H. and Okada. T. S. (1979). Differentiation and dedifferentiationof rat lens epithelial cells in short- and long-term cultures. Dev. Growth Difler. 21. 205-20. Harding. J. J. and Crabbe, M. J. (1984). The Lens: Develoument.Proteins. Metabolismand Cataract. In The Eye.&Vol.1B (Ed. Davson,H.). Pp. 207-440. Academic Press:New York. Jacob,T. J. (1987). HumanLensepithelialcellsin culture : a quantitative evaluation of growth rate and proliferative capacity. Exp. Eye Res. 45, 93-104. Jones,R. E.. DeFeo,D. and Piatigorsky. J. (198 1). I. Lens. Initial studieson cultured embryonic chick lens epithelial cellsinfected with a temperature-sensitiverous sarcomavirus. Vision Res. 2 1, 5-9. La Rosa,G.J. and Gudas,L. J. (1988). An early effect of retinoic acid: cloning of an mRNA (Era-l) exhibiting rapid and protein synthesis-independentinduction during teratocarcinomastemcell differentiation. Prof. Natl. Acad. Sci. U.S.A. 85, 1329-33. Lenstra, J. A., Hukkelhoven, M. W.. Groeneveld. A. A., Smits. R. M., Weterings, P. J. and Bloemendal. H. (1982). Geneexpressionof transformedlenscells.Exp. Eye Res. 35, 549-55.

Lipman.R. D. and Taylor. A. (198 7). The in vitro replicative potential and cellular morphology of human lens epithelialcellsderivedfrom different ageddonors.Cum. Eye Res. 6, 1453-7. Maden, M.. Ong. D. E., Summerbell, D. and Chytill, F. (1988). Spatial distribution of cellular protein binding to retinoic acid in the chick limb bud. Nature 335. 733-5. Maddox, A.-M., and Haddox. M. K. (1988). Characteristics of cyclic AMP enhancementof retinoic induction of increased transglutaminase activity in HL60 cells. Exp. Cell Biof. 56, 49-59. Mascarelli.F.. Courtois,Y. and Arruti. C. (1989). The effect of eye-derived-growth factor (EDGFs)on methionine incorporation in the different cell populationsof bovine adult lensin organ culture. Exp. Eye Res. 48. 177-86.

354 Mavilio, F., Simeone, A., Boncinelli, E. and Andrew& P. W. (1988). Activation of four homeobox gene clusters in human embryonal carcinoma cells induced to differentiate by retinoic acid. Diflerentiation 37, 73-V. Menko, A. S. and Boettiger, D. (1988). Inhibition of chicken embryo lens differentiation and lens junction in culture by pp-60”.s”. Mol. Cell Biol. 8, 1414-20. Menko, A. S., Klukas. K. A., Liu. T.-F., Quade, B., Sas, D. F., Preus, D. M. and Johnson, R. G. (1987). Junctions between lens cells in differentiating cultures ; structure, formation, intercellular permeability and junctional protein expression. Dev. Biol. 123, 307-20. Muggleton-Harris, A. L. and Higbee, N. (1987). Factors modulating mouse lens epithelial cell morphology with differentiation and development of a lentoid structure in vitro. Development 99, 25-32. Nagineni, C. N. and Bhat, S. P. (1989). Human fetal lens epithelial cells in culture : an in vitro model for the study of crystallin expression and lens differentiation. Curt-. Eye Res. 8, 282-91. Patek, C. E. and Clayton, R. M. (1985). A comparison of changing pattern of crystallin expression in vivo and in long-term primary cultures in vitro and in response to a carcinogen. Exp. Eye Res. 40, 357-78. Patek, C. E. and Clayton, R. M. (1986a). Patterns of crystallin expression during differentiation in vitro of five chick genotypes each with different effects on the rate of cell growth. Exp. Eye Res. 43, 1111-26. Patek, C. E. and Clayton, R. M. (1986b). Alterations of crystallin gene expression during subculture of chick lens cells. Exp. Eye Res. 43, 595-606. Patek, C. E. and Clayton, R. M. (1986~). Retinoic acid is lentoidogenic but differentially affects delta-crystallin expression by chick lens cells in vitro. In Coordinated Regulation of Gene Expression (Fds Clayton, R. M. and Truman, D. E. S.). Pp. 377-83. Plenum Press: New York. Patek, C. E. and Clayton, R. M. (1988). The influence of the genotype on the process of ageing of chick lens cells in vitro. Exp. Cell Res. 174, 33043. Patek, C. E., Vornhagen, R., Rink. H. and Clayton, R. M. (1986). Developmental changes in membrane protein expression by chick lens cells in vivo and in vitro and the detection of Main Intrinsic Polypeptide (MIP). Exp. Eye Res. 43, 2940. Piatigorsky, J. (198 1). Lens differentiation in vertebrates : a and molecular features. review of cellular Difjerentiation 19, 1134-53. Piatigorsky, J. (1987). Gene expression and genetic engineering in the lens. Invest, Ophthalmol. Vis. Sci. 28, V-28. Ramaekers, F. C. S., Hukkelhoven, M. W., Groenveld, A. and Bloemendal, H. (1984). Changing protein patterns during lens cell ageing in vitro. Biochim. Biophys, Acta 799, 221-9. Reddan, J. R., Chepelinsky, A. B., Dziedzic. D. C.. Piatigorsky, J. and Goldenberg, E. M. (1986). Retention of lens specificity in long-term cultures of diploid rabbit lens epithelial cells. Diferentiation 33, 168-74. Reddy, V. N., Lin, L. R.. Arita, T., Zigler, J. S. and Huagn. Q. L. (1988). Crystallins and their synthesis in human

C. E. PATEK

AND

R. M. CLAYTON

lens epithelial cells in tissue culture. Exp. Eye Res. 47, 465-78. Richardson, N. A. and McAvoy, J. W. (1988). Age-related changes in fibre differentiation of rat lens epithelial cells in vitro. Exp. Eye Res. 46, 259-67. Rink, H. (1984). Growth potential, repair capacity and protein synthesis in lens epithelial cells during aging in vitro. Monogr. Dev. Biol. 17, 94-107. Rink, H. and Vornhagen. R. (1980). Rat lens epithelial cells in vitro. II. Changes of protein patterns during aging and transformation. In Vitro 16, 277-80. Sherman, M. I. (1986). Retinoids and Diflerentiation. CRC Press: Boca Raton, Fl. Simonneau, L.. Crisanti, P., Lorinet, A. M., Alliot, F.. Courtois, Y., Calothy, G. and Pessac, B. (1986). Crystallin gene expression and lentoid body formation in quail embryo neuroretina cultures transformed by the oncogene retrovirus Mill Hill 2 or Rous Sarcoma Virus. Mol. Cell Biol. 6, 3704-10. Simmoneau, L., Herve, B., Jacquemin. E. and Courtois, Y. (1983a). State of differentiation of bovine epithelial cells in vitro. Modulation of the synthesis and of the polymerisation of specific proteins (crystallins) and nonspecific proteins in relation to cell divisions. Exp. Cell Res. 145, 433-46. Simonneau, L., HervC. B., Jacquemin. E. and Courtois, Y. (1983b). State of differentiation of bovine epithelial lens cells in vitro. Relationship between the variation of the cell shape and the synthesis of crystallins. Cell Dijf. 13. 18 S-90. Skreb, N., Hofman, L. J., Skreb. Y., Suchanek, E. and Sermon, D. (1984). Cyclic nucleotides affect growth and differentiation of cultured rat embryonic shields. Dev. Biol. 101, 367-72. Srivastava, 0. P. and Srivastava, K. (1989). Human lens membrane proteins: Purification and age related distributional changes in the water-soluble and insoluble protein fractions. Exp. Eye Res. 48, 161-75. Thaller, C. and Eichele. G. (1988). Characterisation of retinoid metabolism in the developing chick limb bud. Development 103. 473-83. Thiele, C. J., Deutsch. L. A. and Israel, M. A. (1988). The expression of multiple proto-oncogenes is differently regulated during retinoic acid induced maturation of human neuroblastoma cell lines. Oncogene 3, 281-8. Van Venrooij. W.. Groeneveld. A., Bloemendal, H. and Benedetti, E. (1974). Cultured calf lens epithelium. I. Methods of cultivation and characteristics of the cultures. 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). Normal and conditionally transformed bovine lens epithelial cell lines containing alpha- and gamma-crystallin. Exp. Eye Res. 34,

71-81.

Westphal, H. (1988). Perturbations of lens development in the transgenic mouse. Cell Diff. Dev. 29 (Suppl.), 33-8.

Wiggert, B., Masterson, E. and Coulombre, A. J. (1983). Changes in retinoid binding levels during development of the chicken cornea. Exp. Eye Res. 37, 499-507.