Enhanced expression of cyclin D1 in senescent human fibroblasts

tmdmims of a@ni Mechanisms of Ageing and Development ELSEVIER

andrkvebpment

81 (1995) 139-157

Enhanced expression of cyclin D, in senescent human fibroblasts Junya Fukami”,

‘,

Kumiko Anno”, Kyoko Ueda”, Taijo Takahashib, Toshinori Ide*”

“Department of Cellular and Molecular Biology, Hiroshima University School of Medicine, Kasumi I-2-3, Hiroshima City, Hiroshima 734, Japan ‘Medical and Biological Laboratories Co. Ltd., Marunouchi 3-5-10, Nakaku, Nagoya 460, Japan

Received 14 January 1995; revision received 29 March 1995; accepted 20 April 1995

Abstract When human fibroblast, TIG-1, was growth-stimulated with fetal bovine serum, the induction level of cell cycle-dependent genes was generally much lower in senescent cells than in young counterparts. Exceptionally, the expression level of cyclin D, in senescent cells was constitutively higher than in young cells and further increased after serum stimulation, which was confirmed by Northern and Western blots and immunoprecipitation. This was also true in other human diploid fibroblast lines, TIG-3 and MRC-5. However, cyclin D,-dependent kinase activity was not detected in senescent cells. When sense- or antisense-cyclin D, cDNA driven by p-actin promotor was transfected into young TIG-1 cells, the number of appeared colonies from sense-strand transfected cultures was lower than that from antisense-strandtransfected ones. However, clones expressing cyclin D, at low or undetectable level which were isolated after transfection with antisense-cyclin D, proliferated up to the same division limit as untransfected and sense-strand transfected cells. Four clones of SV40-transformed TIG-1 expressed cyclin D, at moderate levels during their extended proliferative lifespan. It appears that, if the extremely overexpressed cyclin D, could cause an inhibition of cell proliferation at senescent stage, cellular senescence occurs regardless of overexpression of cyclin D,. Keywords: Cyclin D,; Human

fibroblasts;

Senescence

* Corresponding Author, Tel.: + 81 82 257 5290; Fax: + 81 82 257 5294; E-mail: [email protected]. hiroshima-u.ac.jp. ‘Present address: National Institute of Bioscience and Human Technology, Agency of Industrial Science and Technology, Higashi l-1, Tsukuba, Ibaraki 305, Japan. 0047-6374/95/$09.50 0 1995 Elsevier Science Ireland Ltd. All rights reserved SSDI 0047-6374(95)01582-N

140

J. Fukarni et al. ! Mechanisms of Ageing and Development 81 (1995) 139-157

1. Introduction

Normal human diploid fibroblasts have a finite proliferative lifespan in culture [1,2]. At the end of proliferative lifespan, senescent cells remain alive for many months but unable to re-enter S phase in response to growth factors [3]. Previous studies have shown that growth factor-stimulated senescent cells are able to carry out many of the prereplicative events, e.g., binding of growth factor to its receptor [4], activation of receptor tyrosine kinase [4], increased nutrient uptake [5] and expression of c-myc, c-Ha-ras, and p53 [6]. However, several key events necessary for entry into S phase failed to take place after serum stimulation of senescent cells: (1) reduced expression of immediate early gene c-j& [7-91, (2) failure in activation of AP-1 [lo], (3) reduced expression of cyclin A, cyclin B, and Cdc2 [l 11, and (4) inability to phosphorylate RB protein [12,13]. Since an increasing body of evidence suggests that expression of G, cyclin genes and activation of cyclin-dependent kinases is responsible for the phosphorylation of RB protein and regulates process from G, toward S phase [14-211, we examined expression of these genes and activation of Cdk kinases and found that expression of several G, genes was lower in serum-stimulated senescent cells than in young cells. Afshari et al. reported that senescent MRC-5 cells expressed no mRNA of cyclin A and cdk2 and at reduced level mRNAs of cyclins C, D,, and E [22]. Won et al. also reported low expression of cyclins D, and D, in senescent Hs68 cells [23]. During the course of the experiments, we also found that, as an exception, the expression level of cyclin D, was constitutively much higher in senescent cells than in young cells. Recent report demonstrated that the expression of cdk2 and cdk4 was repressed but that of cyclin D, and E was elevated in senescent WI-38 cells 1241. Dulic et al. also reported using IMR-90 overexpression of cyclin D, and E, and formation of enzymatically inactive complexes of cyclin E-Cdk2 and cyclin D,-Cdk2 [25]. Therefore, altered expression of cyclin D, in senescent cells was a cell line-dependent phenomenon. The present paper reports that senescent human fibroblast, TIG-1, expresses cyclin D, at high level whose product forms enzymatically inactive complexes and suggest that high expression of cyclin D, is not a causal event for growth repression. 2. Materials and methods 2.1. Cell culture Human diploid fibroblasts, TIG-1, were cultured as described previously [3]. Though the reduction of growth potential was generally observed in accordance with the increase in division age, the growth rate of TIG-1 cells estimated from the routine schedule of subculture, every 3-4 days by a 1:4 split ratio, did not change markedly up to 50 PDL. After 55 PDL reduction of growth rate and morphological changes began to occur markedly, and the cells usually could not proliferate over 60 PDL. Senescent cells were used at least 3 weeks after the

J. Fukami et al. / Mechanisms of Ageing and Development 81 (1995) 139-157

subconfluent cells did not proliferate serum as previously reported [26].

any further

in medium containing

141

10%

2.2. Assay of DNA synthesis The cells plated in 60-mm dishes were labeled with [3H]thymidine (37 kBq/ml) for 24 h. After labeling, the cells were solubilized in STE buffer (0.6% SDS, 5 mM 2NaEDTA, 150 mM NaCl, 50 mM Tris-HCl, pH 7.5), and DNA was trapped on glass fiber filters (GF/c; Whatman). Radioactivity was counted by a liquid scintillation spectrometer. Assay of DNA synthesis of cells plated on coverslips was done as previously reported [27]. 2.3. Northern blot analysis Isolation and Northern blot analysis of cellular RNA were done as previously reported [5]. The equal amount of RNA on each lane was confirmed by ethidium bromide staining after each run. Plasmids used were as follows: C3sp6v-fos carrying viral oncogene v-fos sequence [28] was a gift from Dr. S.R. Rittling, pMYC54 ca.rrying human c-myc first exon sequence [29] was a gift from Dr. Dalla-Favera, pcdk2-H carrying human cdk2 cDNA sequence [30] was a gift from Dr. E. Hara, pGEM3-HCYC-C carrying human cyclin C cDNA sequence [14] was a gift from Mr. N. Tsuyama, pCB204.1 carrying mouse “cyclin D, cDNA sequence [16] was a gift from Dr. C.J. Sherr, pCYCDl-H123 carrying human cyclin D, cDNA sequence [18] was a gift from Dr. Y. Xiong, pHcyclinA carrying human cyclin A cDNA sequence [31] was a gift from Dr. H. Yasuda, and pHcyclinB carrying human cyclin B cDNA sequence [32] was a gift from Dr. H. Yasuda. 2.4, Immunojluorescent staining Cells on glass coverslips were fixed with 10% formalin and with cold acetone and stained as described [33] with anti-cyclin D, antibody and with fluoresceinconjugated rabbit anti-rabbit IgG (MBL, Nagoya, Japan). 2.5. Irnmunoprecipitation with anti-cyclin D, antibody Cells in a 60-mm dish were washed with methionine-free medium, and pre-incubated at 37°C in methionine-free medium for 30 min. Then the cells were labeled with 3.7 MBq/ml [35S]methionine (Tran 35S-label METABOLIC LABELING REAGENT, 370 TBq/mmol; ICN) for 3 h. The labeled cells were suspended in lysis buffer (RPM1 buffer [50 mM Tris-HCl, pH 7.8, 150 mM NaCl, 20 mM EDTA, and 0.5% Nonidet P-401 containing 200 PM Na,VO,, 21 pg/ml aprotinin, 100 pug/ml phenylmethylsulfonyl fluoride, and 10 pg/ml leupeptin) and left on ice for 15 min. In some experiments, EBC buffer (50 mM Tris-HCl, pH 8.0, 120 mM NaCl, 1% Nonidet P-40) containing 200 PM Na,VO,, 21 pg/ml aprotinin, 100 pg/ml PMSF, and 10 pug/ml leupeptin was used. Lysates were clarified by centrifugation at 9000 x g for 10 min at 4”C, and the supernatants were mixed with protein A-agarose beads (KURABO, Osaka) and stood for 30 min at 4°C. Then they were

142

J. Fukami

et al. 1 Mechanisms

of Age&g and Development

81 (1995) 139- 157

clarified by centrifugation at 9000 x g for 1 min at 4°C. The supernatants were reacted with anti-cyclin D, antibody for 1 h at 4°C then mixed with protein A-agarose beads for 1 h at 4°C. Immunocomplexes on beads were precipitated as above and were washed with lysis buffer 4 times and with washing solution (10 mM Tris-HCl, pH 7.6, 0.1% Nonidet P-40) once. The immunocomplexes were resuspended in 10 ~1 water and added to 5 ~1 Laemmli’s buffer (312.5 mM Tris-HCl, pH 6.8, 10% SDS, 50% glycerol, 25% 2-mercaptoethanol, 0.005% BPB), and boiled at 100°C for 5 min. After centrifugation at 9000 x g for 5 min, the supernatants were run on SDS-polyacrylamide gel electrophoresis [34]. Electrophoresed gels were treated with EN3HANCE (NEN) and processed for fluorography. 2.6. Protein kinase assay of immunoprecipitant Antigen-antibody complexes on protein A-agarose beads were prepared as above, except that the cells were not radiolabeled. The same amount of proteins in each sample was reacted with antibody. The complexes were suspended in 40 ,~l kinase buffer (20 mM Tris-HCl, pH 7.4, 10 mM MgCl,, 1 mM EGTA, 4.5 mM 2-mercaptoethanol) and started the reaction by adding histone Hl (Boehringer Mannheim), 20 PM ATP, and 185 kBq [y-32P]ATP (220 TBq/mmol; Amersham). After incubation for 20 min at 30°C the samples were added to 12.5 ~1 Laemmli’s buffer, boiled at 100°C for 5 min and electrophoresed on SDS-polyacrylamide gel. Phosphorylated proteins were visualized by autoradiography of dried slab gels. 2.7. Western blot analysis Cells (1 x 106) were suspended in EBC buffer containing 200 ,uM Na,VO,, 21 pug/ml aprotinin, 100 pg/ml PMSF, and 10 pg/ml leupeptin and left on ice for 15 min. Cell lysates (20 ~1) containing equal amount of proteins were mixed with 5 ~1 Laemmli’s buffer, boiled at 1OO’C for 5 min, and applied on SDS-polyacrylamide gel. After electrophoresis, proteins were transferred to blotting membrane using Electroblot Apparatus (ATTO). The membrane was soaked in blocking buffer (TBS-T (50 mM Tris-HCl, pH 7.5, 0.45% NaCl, 0.05% Tween 20) containing 5% skim milk) for 1 h at room temperature and incubated in blocking buffer containing anti-cyclin D, antibodies at room temperature overnight. The filter was then washed with TBS-T 5 times and incubated in blocking buffer containing horseradish peroxidase-anti-IgG conjugate (Amersham) for 1 h at room temperature. The filter was washed with TBS-T 5 times, reacted with detection reagent (NEN) for 1 min at room temperature, and the reaction products were visualized by fluorography. 2.8. Transfec tion cDNA encoding human cyclin D, [18] was subcloned into the pML-neo vector (MMTV-LTR promoter-driven cyclin D,) [35], pMIK-hygB (SRc( promoter-driven cyclin D1) [36], and pH/3APr-lneo (p-actin promoter-driven cyclin D,) [37]. TIG-1 cells (2 x 106) suspended in 200 ml PBS( + ) (8.0 g/l NaC1, 0.2 g/l KCI, 0.1 g/l CaCl,, 0.06 g/l MgS0,7H,O, 0.1 g/l MgC1,6H,O, 1.44 g/l Na,HP0,2H,O, 0.2 g/l KH,PO,) were transfected with plasmid DNA (20 pg) by electroporation using

J. Fukami et al. / Mechanisms

of Ageing and Development

143

81 (1995) 139-157

Gene Pulser (Bio-Rad) at 0.8 kV and 25 ,uF [38]. They were plated in DMEM containing 10% FBS and incubated at 37°C for the first 48 h, and then selected for 2 weeks in medium with 400 pg/ml G418 (Gibco) or 100 pgg/ml hygromycine B. Transcription of cyclin D, cDNA subcloned in pML-neo was controlled by treatment with 5 ,uM dexamethasone. Appeared colonies were counted after fixing with 10% formalin and with ethanol and staining with Giemsa solution. 3. Results 3.1. Northern analysis of cell cycle dependent genes Senescent TIG-1 cells used here were prepared after they ceased proliferation at around 60 PDL keeping them for more than 3 weeks with medium changes twice a week. They did not initiate DNA synthesis after medium change (Fig. 1). When young TIG-1 cells were growth-arrested by culturing them in serum-free medium for 2 days and then stimulated with serum, stimulation of DNA synthesis /Fis. 1) and induction of cell cycle-dependent genes (Fig. 2) were demonstrated. A!though

-Q-young cells + senescent cells

G

c<

0

0

10 Time after

20 30 stimulation (h)

’ i

40

Fig. I. Induction of DNA synthesis in young and senescent TIG-I cells. Go-arrested young (33 PDL) and senescent (60 PDL) TIG-1 cells plated in 24-well multiplate were stimulated to proliferated with 10% FBS. They were labeled continuously with [3H]thymidine (37 kBq/ml) for indicated periods.

144

J. Fukami

et al. / Mechanisms

qf Age&

and Development

81 (1995) 139-157

PCNA+

TK-

c

histone H3-

Fig. 2. Expression of cell cycle-related genes in young and senescent cells. Go-arrested young (33 PDL) and senescent (60 PDL) TIC-I cells were stimulated with 10% FBS. Cytoplasmic RNA was extracted at indicated time and the steady state level of mRNA was examined by Northern blot analysis.

previous reports indicated that among primarily inducible genes c-fos did not express in senescent cells [779], senescent TIG-1 cells induced c-fos although at lower extent (Fig. 2). Expression of cdk2, cyclin C, PCNA, and cyclin A, which occurred in mid/late Gi phase in young cells, was also observed in senescent cells although at lower level. Induction of such genes of S phase marker as histone H3 and thymidine kinase and such G,/M marker gene as cyclin B was absent in senescent TIG- 1 cells. Expression level of cyclin D, was much higher in unstimulated senescent cells than in young cells and it enhanced after serum stimulation (Fig. 2). This observa-

J. Fukami et al. / Mechanisms of Ageing and Development 81 (1995) 139- I57

145

tion is, consistent with the recent results from WI-38 [24], IMR-90 [25], and from TIG-3 cells (our unpublished data), but is different from results using MRC-5 cells [22] and Hs68 cells [23] in which expression level of cyclin D, is not high in senescent cells and increased only at reduced extent after serum stimulation as compared to young cells. Therefore, altered expression of cyclin D, in senescent cells was a cell line dependent-phenomenon. Does expressional change in cyclin D, occur at the very last stage of cellular senescence? When TIG-1 cells at late passage (53 PDL) were stimulated with serum, they still induced DNA synthesis and histone H3 expression at reduced level than young cells at 39 PDL, but did c-myc expression at the same level as young cells did (Figs. 3 and 4). Expression level of cyclin D, in TIG-1 cells at 53 PDL was already much higher than that in cells at 39 PDL (Fig. 4). Among cell cycle-dependent genes tested, all but cyclin D, were more or less down-regulated in senescent TIG-1 cells and we focused the further studies on the expression of cyclin D,. 3.2. Expression of cyclin D, protein in senescent cells Higher expression of cyclin D, in senescent cells was also confirmed by W,estern blot (Fig. 5A), by immunoprecipitation of cellular proteins after labeling with

MO5 stimulated non-stimulated

39PDL

5OPDL

53PDL

56PDL

Fig. 3. lnduction of DNA synthesis in late-passage TIC-I cells. C&-arrested TIG-I cells plated in 60-mm dishes were stimulated to proliferated with 10% FBS. They were labeled with [‘Hlthymidine (37 kBq/ml) for 24 h and inCorporation of [‘Hlthymidine into DNA was measured.

146

J. Fukami et al. / Mechanisms

ofAgeing and Development 81 (1995) 139-157

39PDL GO 0.5 -m--

2

3

GO 0.5 -24 I B--m-

2

3

24h

histone H3-

Fig. 4. Expression levels of cell cycle-related genes in young and late-passage TIC&I cells were stimulated with 10% FBS. Cytoplasmic RNA was extracted steady state level of mRNA was measured by Northern blot analysis.

TIC-I cells. Go-arrested at indicated time and the

[35S]methionine (Fig. 5B), and by immunofluorescent staining (Fig. 6). By longer exposure of the gel, several additional protein were observed (data not shown) and were similar to those reported by others previously [39]. Increased expression of cyclin D, in senescent cells was also observed in 2 other human fibroblast lines, TIG-3 and MRC-5 (Fig. 7A), while the reason why others reported senescent MRC-5 did not express cyclin D, at high level [22] was unclear. Different anti-cyclin D, antibodies also indicated higher expression of cyclin D, in senescent cells (Fig. 7B). An additional band was sometimes observed just below the band of cyclin D, but it was not characterized further (Figs. 5B and 7). In the present study, the cells were lysed and reacted with antibodies in a buffer solution, RPMI. When another buffer solution, EBC buffer, was used, the protein bands coprecipitated with anti-cyclin D, antibodies showed quantitatively different pattern between young and senescent cells (Fig. 8). In young cells protein bands with higher molecular weight were dominant whereas in senescent cells proteins with lower molecular weight were obvious. It was also observed when different anti-cyclin D, antibodies were used, while intensity of each band varied (Fig. 7B). Although we could not conclude that all these proteins formed complex with cyclin D, protein, most of these protein bands were undetectable by Western blot analysis

J. Fukami et al. / Mechanisms of Ageing and Development 81 (1995) 139-157

147

and by immunoprecipitation with control serum. Some of these proteins may correspond to the reported inhibitor against kinase activity of cyclin/Cdk complex [64-671, and may contribute to the functional differences of cyclin D, between young and senescent cells. 3.3. Kinase activities in immunoprecipitants The same amount of proteins extracted from either young or senescent cells was used for immunoprecipitation with anti-cyclin D, antibody and the resulting immunoprecipitant was used for its kinase activity using histone Hl as a substrate. Kinase activity of immunoprecipitants from young cells increased sharply 12 h after serum stimulation, whereas it was negligible from senescent cells (Fig. 9). It was of interest that phosphorylation of a protein, possibly cyclin D, as estimated from molecular size, was high in samples from young cells but was negligible in those from senescent cells.

normal rabbit

anti-cyclin Dl antibody

I

serum

97.4kD66.2 -

cyclin D1-w

Fig. 5. Detection of cyclin D, protein in young and senescent TIC-I cells. (A) Whole cellular proteins were extracted from young (35 PDL) and senescent (60 PDL) TIG-I cells and cyclin D, was detected by Western blot analysis. (B) Cell extracts from [?S]methionine-labeled young (32 PDL) and senescent (60 PDL) TIG-I cells were immunoprecipitated with anti-cyclin D, antibody (pAK) or normal rabitt serum in RPM1 buffer.

48

J. Fukami

et ul. / Mechanisms

of' Ageing

and Druelopment

81 (1995)

139-157

J. Fukami et al. / Mechanisms of Ageing and Development 81 (1995) 139-157 TIG-3

149

1MRC-5

I

Yo”“g pAK

--25m

senescent

C-l pAK

--25lll C-I m

97.4kD

07.4kD--

-

66.2

-

45.0

-

4.5.0 -

e

cyclin Ill-+

cyclin Dl d 31.0

_ A

-

-

-v--

+

--

cyclin D

.31.0-

-

21.5

Bcp.m * ) 9-m il-*r

66.2 -

21..s-

-

B

_”

Fig. 7. Immunoprecipitation of cyclin D, protein. (A) Young (35 PDL) and senescent (84 PDL) TIG-3 and young (35 PDL) and senescent (64 PDL) MRC-5 cells were labeled with [‘sS]methionine and the lysates were immunoprecipitated with anti-cyclin D, antibody (pAK) in RPM1 buffer. (B) Young (32 PDL) and senescent (60 PDL) TIG-1 cells were labeled with [%]methionine and the lysates were immunoprecipitated in RPM1 buffer with different anti-cyclin D, antibodies. pAK. polyclonal antibody against GST-cyclin D, proteins. 25m, polyclonal antibody against peptide DFIEHFLSKMPEAEENKQIIRKHAQ (207-231 a.a. in cyclin D, protein). c-t, polyclonal antibody against c-terminal peptide of cyclin D, protein.

3.4. Transfection of cyclin D, cDNA

To see whether high expression of cyclin D, has a causal role for repression of proliferation in senescent cells, young TIG-I cells (30 PDLs) were transfected with plasmid DNA which carried cyclin D, cDNA under control of @-actin promoter and drug resistant gene, and assayed growth of transfectants by colony formation in the selection medium. Colony-forming efficiency of sense-strand cyclin D, transfected culture was lower than that of antisense-strand transfected one (Table 1). Pooled surviving colonies from sense-strand transfected cultures did not show overexpression of cyclin D, protein (data not shown). Extremely high expression of cyclin D, may be harmful for cell growth and therefore highly expressing cells were not cloned. However, experiments by using vector containing different promotor (SRa) gave no significant difference between sense- and antisense-strand transfected cultures. Young TIG-3 cells were also transfected with plasmid DNA carrying MMTV promotor-driven cyclin D, sequence (sense or antisense) to examine relationship between expression level of cyclin D, and division limit. Clones were selected by G418 resistance in the absence of dexamethasone and the isolated clones were further cultured either in the presence or absence of dexamethasone (5 /rmol/ml).

150

J. Fukami et al. 1 Mechmisms

of Age@

and Development

81 (1995) 139-157

The PDL of these ceils was estimated from the approximate number of divisions required for the assumed single transfected cells to reach a confluent monolayer in 60-mm dishes where the cell number was examined. We isolated 23 clones from sense-strand-transfected cultures and 42 clones from antisense-strand:transfected cultures. The expression level of cyclin D, determined by immunoprecipitation was unaltered in most clones by the presence of dexamethasone and was found to vary among isolated clones: some sense-strand-transfected clones showed slightly higher expression level of cyclin D, than, but others showed the same level as, untransfected control, and some antisense-strand-transfected clones showed no expression of cyclin D,, but others showed the same level as, untransfected control (some examples were shown in Fig. 10). However, all these clones proliferated up to around 60 PDLs ( + 5 PDLs), indicating that amount of cyclin D,, unless extremely overexpressed, appeared to be nothing to do with cellular senescence. 3.5. Expression of cyclin D, in SV40 T-antigen transformed cells Some genes whose expression increased with increasing division age of normal fibroblasts also expressed at high level during extended proliferative lifespan of SV40-transformed cells [40,41]. We next examined expression level of cyclin D, in SV40-transformed cells to see whether the expression of cyclin D, also increases in accordance with increasing division age of transformed cells. Normal TIG-I cells

The amount of anti-cyclin Dl --1.5

y0llng 2 - 3

senescent - 1.5 - 2 -3pl

Fig. 8. Electrophoretic pattern of proteins co-immunoprecipitdted with anti-cyclin D, antibody. [35S]methionine-labeled young (32 PDL) and senescent (60 PDL) TIG-1 cells were lysed and immunoprecipitated with anti-cyclin D, antibody (pAK) in EBC buffer instead of RPM1 buffer.

J. Fukami ef al. / Mechanisms of Ageing and Development 81 (1995) 139-157

senescent

wung

GB 2 _-_--_---w

97.4kD

-

66.2

-

histone 21.5

6

151

12

24

GB

2

6

12 24hr

H1-w -

Fig. 9. Cyclin D,-dependent kinase activity in young and senescent TIG-1 cells. Go-arrested young (32 PDL) and senescent (60 PDL) TIC-1 cells were stimulated with 10% FBS for indicated period. Immunoprecipitants with anti-cyclin D, antibody (pAK) were assayed for kinase activity using histone HI as a substrate.

were transfected with plasmid pMT-1OD which carries SV40 T-antigen as reported previously [42]. Transformed clones were isolated and propagated [43]. When they were ready to use for experiments, they were already close to the end of the lifespan of normal cells. It should be considered that the estimated PDL of transformed cells was always underestimated because of relatively extensive cell loss during cloning Table 1 Colony formation of TIG-I cells after transfection with sense- or antisense-cyclin D, cDNA Plasmids

Colony forming ability Vector

pHbA-D,” PMIK-D,~ PMMTV-D,~ dex( - ) dex( + )

39 x 10-S 35 x 10-s

Sense

Antisense

5.8 X IO-5 1.6 x 1O-5

21.5 x lop5 1.8 X 10-s

43 x 10-5 32 x 1O-5

12 X 10-S 19 x 10-5

“Plasmid containing p-actin promotor-driven sense- or antisense-cyclin ( 106) were selected in 400 pg/ml G418 for 2 weeks. bPlasmid containing SR-1 promotor-driven sense- or antisense-cyclin (8 x 105) were selected in 100 pg/ml hygromycin B. ‘Plasmid containing MMTV promotor-driven sense- or antisense-cyclin (8 x 105) were selected in 400 pg/ml G418 in the absence or presence of

D, strand. Transfected cells D, strand. Transfected cells D, strand. Transfected cells 5pM dexamethdsone.

clone

Al-l

AZ-4

AZ-18

deu.

I

t

i

qrlin

s-n

I

AZ-7

t

AZ-14

s-13

S-II

I

t

I

DI

Fig. 10. lmmunoprecipitation of cyclin D, protein in TIG-I clones transfected with sense-strand or antisense-strand of cyclin D, cDNA. TIG-I clones transfected with sense-strand (S-clones) or antisensestrand (A-clones) of cyclin D, cDNA were isolated in the absence of dexamethasone and cultured in the presence (+ ) or absence ( - ) of dexamethasone (5 pmol/ml). They were labeled with [‘sS]methionine for 3 h at around 50-55 PDLs and the lysates were immunoprecipitated with anti-cyclin D, antibody (pAK) in RPM1 buffer.

process. Therefore, we could not use young transformed cells. As shown in Fig. 11, expression of cyclin D, was much lower in several SV40-transformed clones than in normal senescent cells. This was confirmed by Northern blot and Western blot analyses (data not shown). 4. Discussion The present paper reports that the expression level of cyclin D, increased when human diploid fibroblasts, TIG-1, TIG-3, and MRC-5, senesced. This observation is consistent with the previous results from WI-38 [24] and IMR-90 [25], but is different from results using MRC-5 [22] and Hs68 [23] in which expression level of cyclin D, is not high in senescent cells and increased only at reduced extent after serum stimulation as compared to young cells. Although we do not know the reason why our results on MRC-5 were inconsistent with others [22], it is likely that altered expression of cyclin D, in senescent cells was a cell line-dependent phenomenon. Accumulating evidence indicated that senescent cells express a dominant inhibitor of cell proliferation which probably controls the process of cellular senescence. This inhibitory activity was demonstrated by cell fusion between young and senescent cells [44,45], by cybrid formation between senescent cytoplast and young cells [46,47], addition of cell membrane fraction from senescent cells to culture medium of young cells [48], or by microinjection of mRNA from senescent cells into young cells [49]. Microcell-mediated chromosome transfer experiments have provided evidence for dominant senescence genes on several distinct chromosomes [50-541. cDNA clones which expressed with increasing division age or expressed at high level in senescent cells were isolated expecting their ability to inhibit cell proliferation [40,55-601. However, cyclin D, has not been listed in these cDNAs. On the other hand, cell cycle-dependent genes which failed to be induced or activated after growth stimulation in senescent cells were searched expecting to explain inability of proliferation. Senescent cells fail to induce some of immediate early and middle/late G, genes after serum stimulation [6-l 11, and they fail to phosphorylate RB protein [ 12,131. Several cell cycle-dependent genes, however, expressed at the same levels as in stimulated young cells [6]. Since cyclin D, is known to be a positive regulator of the cell cycle [16,61] and expresses at high level

J. Fukami et al. / Mechanisms of Ageing and Development 81 (1995) 139-157

153

in some types of tumor cell [17,62,63], no kinase activity of the cyclin D, complex in senescent cells is consistent with the inability of cell proliferation [25]. Recently, several low-molecular-weight proteins, such as p16 [64], p21 [65], p27 [66], and p28 [67], were isolated which bound and inhibited kinase activity of cyclin/Cdk complexes resulting in growth inhibition. One of these, ~21, was isolated as a senescence-dependent inhibitor of DNA synthesis [68] and actually increased its expression when normal and SV40-transformed human fibroblasts senesced [42]. Proteins which coprecipitate with anti-cyclin D, antibodies are different between young and senescent cells such that in senescent cells several protein bands of low molecular weight appeared to be preferentially associated with cyclin D, protein (Fig. 8).

TIG-1

I

tfansfoImants

m

cyclin Dl + 31.0 -

Fig. 11. Immunoprecipitation of cyclin D, protein in SV40-transformed TIC-I cells. Normal young (32 PDL) and senescent (60 PDL) TIC&I and SV40-transformed clones IOD-I (70 PDL), lOD2-8 (67 PDL), 20PEG1-4 (65 PDL) and 20PEG8-13 (63 PDL) were labeled with [%]methionine for 3 h and the lysates were immunoprecipitated with anti-cyclin D, antibody (pAK) in RPM1 buffer.

154

J. Fukumi et nl. / Mechanisms

of Ageing and Development

81 (1995) 139- 157

However, it is unexpected that senescent cells expressed cyclin D, at high level and further induced it after serum stimulation. The question is whether age-dependent overexpression of cyclin D, is something to do with cellular senescence and whether it is a result of cellular ageing or a cause of ageing. Although we could not directly answer these questions, colony-forming efficiency of /I-actin promotordriven sense-strand cyclin D, transfected culture was lower than that of antisensestrand transfected one (Table 1). Pooled surviving colonies from sense-strand (either B-actin or MMTV-promotor-driven cyclin Dr) transfected cultures did not overexpress cyclin D, protein. Taken together previous results reporting that low colony-forming efficiency in cyclin D, transfected culture [69] and cyclin D,-mediated inhibition of replicative DNA synthesis [70], extremely high expression of cyclin D, is likely to be harmful for cell growth. Some clones obtained from antisense-strand-transfected cultures failed to express detectable cyclin D, protein but they proliferated to the same PDLs as untransfected control cells and sensestrand transfected clones. Therefore, it is obvious that, if the extremely overexpressed cyclin D, could cause an inhibition of cell proliferation at senescent stage, cellular senescence itself occurs regardless of overexpression of cyclin D,. SV40-transformed cells continued proliferation over limitation of PDLs of normal cells and these transformed cells expressed cyclin D, at much lower level than normal senescent cells (Fig. 11). This was shown in 4 examples of SV40-transformed clones with no exception. These results may be explained by 2 possible ways: (1) SV40 T-antigen blocked extremely high expression of cyclin D, and (2) SV40-transformed cells expressing cyclin D, at moderate level were selected from the mixed population to proliferate over PDLs of the limit of normal cells. Whatever the mechanisms are, the results appear to be consistent with the explanation that the expression of cyclin D, at extremely high level is not favourable for cells to continue proliferation.

References [l] L. Hayflick and P.S. Moorhead, The serial cultivation of human diploid cell strains. Exp. Cell Res., 25 (1961) 585-621. [2] L. Hayflick, The limited in vitro lifespan of human diploid cell strains. Exp. Cell Rex, 37 (1965) 614-636. [3] Y. Tsuji, T. Ide. S. lshibashi and K. Nishikawa, Loss of responsiveness in senescent human TIG-1 cells to the DNA synthesis-inducing effect of various growth factors. Me&. Ageing Deu., 29 (1984) 219-232. [4] G.S. Gerhard, P.D. Phillips and V.J. Cristofalo, EGF- and PDGF-stimulated phosphorylation in young and senescent WI-38 cells. E.up. Cell Rex, 193 (1991) 87-92. [5] F. Kihara, Y. Tsuji, M. Miura, S. Ishibashi and T. Ide, Events blocked in prereplicative phase in senescent human diploid cells, TIG-I. following serum stimulation. Mech. Ageing Dev., 37 (1986) 103-l 17. [6] S.R. Rittling, K.M. Brooks, V.J. Cristofalo and R. Baserga, Expression of cell cycle-dependent genes in young and senescent WI-38 fibroblasts. Proc. Natl. Acad. Sci. USA, 83 (1986) 3316-3320. [7] Y. Paulsson, M. Bywater. S. Pfeifer-Ohlsson., R. Ohlsson, S. Nilsson. C.-H. Heldin, B. Westermark and C. Betsholtz, Growth factors induce early pre-replicative changes in senescent human fibroblasts. EMBO J., 5 (1986) 2157-2162.

J. Fukami et al. / Mechanisms

of Ageing and Development

81 (1995) 139-157

155

[8] P.D. Phillips, R.J. Pignolo, K. Nishikura and V.J. Cristofalo, Renewed DNA synthesis in senescent WI-38 cells by expression of an inducible chimeric c-fos construct. J. Cell Physiol., 151 (1992) 206-212. [9] T. :Seshadri and J. Campisi, Repression

of c-fos transcription and an altered genetic program in senescent human fibroblasts. Science, 247 (1990) 205-209. [IO] K. Riabowol, J. Schiff and M.Z. Gilman, Transcription factor AP-1 activity is required for initiation of DNA synthesis and is lost during cellular aging. Proc. Natl. Acad. Sci. USA, 89 (1992) 157-161. [It] G.H. Stein, L.F. Drullinger, R.S. Robetorye,

O.M. Pereira-Smith and J.R. Smith, Senescent cells fail to express cdc2, cycA and cycB in response to mitogen stimulation. Proc. Natl. Acad. Sci. USA, 88 (1991) 11012-l 1016. [12] G.H. Stein, M. Beeson and L. Gordon, Failure to phosphorylate the retinoblastoma gene product in senescent human fibroblasts. Science, 249 (1990) 666-669. [l3] P.A.. Fureal and C. Barrett, Failure of senescent cells to phosphorylate RB protein. Oncogene, 6 (19131)110991113. [14] D.J. Lew. V. Dulic and S.I. Reed, Isolation of three novel human cyclins by rescue of G, cyclins (cm) function in yeast. Cell. 66 (1991) 1197-1206. [15] A. Koff, F. Cross, A. Fisher, J. Schumacher, K. Leguellec, M. Philippe and J.M. Roberts, Human cyclin E, a new cyclin that interacts with two members of the CDC2 gene family. Cell 66, (1991) 1217-1228. [l6] H. Matsushime, M.F. Roussel, R.A. Ashmun and C.J. Sherr, Colony-stimulating factor 1 regulates novel cyclins during the G, phase of the cell cycle. Cell, 65 (1991) 701-713. [I71 T. Motokura, T. Bloom, Y.G. Kim, H. Jueppner, J. Ruderman, H. Kronenberg and A. Arnold, A novel cyclin encoded by a bcl l-linked candidate oncogene. Nature, 350 (1991) 512-515. [I81 Y. Xiong, T. Connolly, B. Futcher and D. Beach, Human D-type cyclins. Cell, 65 (1991) 691-699. [I91 V. Dulic, E. Lees and S.I. Reed, Association of human cyclin E with a periodic G,-S phase protein kinase. Science, 257 (1992) 19581961. [20] A. Koff, A. Giordano, D. Desai, K. Yamashita, W. Harper, S. Elledge, T. Nishimoto, D.O. Morgan, R. Franza and J.M. Roberts, Formation and activation of a cyclin E-cdk2 complex during the G, phase of the human cell cycle. Science, 257 (1992) 1689-1693. [21] E. Lees, B. Faha, V. Dulic, S.I. Reed and E. Harlow, Cyclin E/cdkZ and cyclin A/cdkZ kinases associates with ~107 and E2F in a temporally distinct manner. Genes Dev., 5 (1992) 187441885. [22] CA. Afshari, P.J. Vojta, L.A. Annab, P.A. Futreal, T.B. Willard and J.C. Barrett, Investigation of the role of G,/S cell cycle mediators in cellular senescence. Exp. Cell Res., 209 (1993) 231-237. [23] K.-A. Won, Y. Xiong. D. Beach and M.Z. Gilman, Growth-regulated expression of D-type cyclin genes in human diploid fibroblasts. Proc. Natl. Acad. Sci. USA, 89 (1992) 9910-9914. [24] F.C. Lucibello, A. Sewing, S. Brusselbach, C. Burger and R. Muller, Deregulation of cyclin D, and E a.nd suppression of cdk2 and cdk4 in senescent human fibroblasts. J. Cell Sci., 105 (1993) 123-133. [25] V. Dulic, L.F. Drullinger, E. Lees, S.I. Reed and G.H. Stein, Altered regulation of G, cyclins in senescent human diploid fibroblasts: accumulation of inactive cyclin E-Cdk2 and cyclin D,-cdk2 complexes. Proc. Natl. Acad. Sci. USA, 90 (1993) 11034-11038. [26] T. Ide, Y. Tsuji, S. Ishibashi and Y. Mitui, Reinitiation of host DNA synthesis in senescent human diploid cells by infection with simian virus 40. E.rp. Cell Res., 143 (1983) 343-349. [27] K. Tsuji, A. Ueno and T. Ide, Dual elfect of protein kinase C on the induction of DNA synthesis by colcemid in G,-arrested human diploid fibroblasts. Cell Struct. Funct., 16 (1991) 73-80. [28] M.E. Greeberg and E.B. Ziff, Stimulation of 3T3 cells induces transcription of c$os proto-oncogenes. Nature, 311 (1984) 433-438. [29] L.W. Stanton, R. Watt and K.B. Marcu, Molecular cloning of gene sequences regulated by platelet-derived growth factor. Cell, 33 (1983) 9399947. [30] J. Ninomiya-Tsuji. S. Nomoto, H. Yasuda, S.I. Reed and K. Matsumoto, Cloning of a human cDNA encoding a CDCZ-related kinase by complementation of budding yeast cdc28 mutation. Proc. Natl. Acad. Sri. USA, 88 (1991) 9006&9010.

156

J. Fukami

et al. i Mechanisms

qf Ageing and Development

81 (1995) 139-157

[31] J. Pines and T. Hunter, Human cyclin A is adenovirus EIA-associated protein p60 and behaves differently from cyclin B. Nature, 346 (1990) 760-763. [32] M.J. Lohka, Mitotic control by metaphase-promoting factor and cdc proteins. J. Cell Sci., 92 (1989) 131-135. [33] Y. Tsuji, T. lde and S. Ishibashi. Correlation between the presence of T-antigen and the reinitiation of host DNA synthesis in senescent human diploid fibroblasts after SV40 infection. Exp. Cell Res., 144 (1983) 165-169. [34] U.K. Laemmli, Cleavage of structual proteins during the assembly of the head of bacteriophage T4. Nature, 227 (1970) 680-685. [35] F. Lee, R. Mulligan, P. Berg and G. Ringold, Glucocorticoids regulate expression of dihydrofalate reductase cDNA in mouse mammary tumor virus chimaeric plasmids. Nature, 294 (1981) 228-232. [36] Y. Takebe, M. Seiki, J. Fujisawa, P. Hoy, K. Yokota, K. Arai. M. Yoshida and N. Arai, SRct promotor: an efficient and versatile mammalian cDNA expression system composed of simian virus 40 early promotor and the R-U5 segment of human T-cell leukemia virus type I long terminal repeat. Mol. Cell. Biol., 8 (1988) 466-472. [37] P. Gunning, J. Leavitt, G. Muscat, S. Ng and L. Kedes, A human b-actin expression vector system directs high-level accumulation of antisense transcripts. Proc. Natl. Acad. Sci. USA, 84 (1987) 4831-4835. [38] T.K. Wong and E. Neumann, Electric field mediated gene transfer. Biochem. Biophys. Res. Commun., 107 (1982) 584-587. [39] Y. Xiong, H. Zhang and D. Beach, D type cyclins associate with multiple protein kinases and the DNA replication and repair factor PCNA. Cell, 71 (1992) 505-514. [40] H. Tahara, E. Hara, N. Tsuyama, K. Oda and T. Ide, Preparation of a subtractive cDNA library enriched in cDNAs which expressed at a high level in cultured senescent human fibroblasts. Biochem. Biophys. Res. Commun., 199 (1994) 1108-I 112. [41] H. Tahara, E. Sate, A. Noda and T. Ide, Increase in expression level of p2lsdil/cipl/wafI with increasing division age in both normal and SV40-transformed human fibroblasts. Oncogene, IO (1995) 835-840. [42] N. Tsuyama, M. Miura, M. Kitahira, S. Ishibashi and T. Ide, SV40 T-antigen is required for maintainance of immortal growth in SV40-transformed human fibroblasts. Cell Struct. Funct., 16 (1991) 55-62. [43] T. Ide, Y. Tsuji, T. Nakashima and S. Ishibashi, Progress of aging in human diploid cells transformed with a tsA mutant of simian virus 40. E.up. Cell Res.. 150 (1984) 321-328. [44] T.H. Norwood, W.R. Pendergrass, C.A. Sprague and G.M. Martin. Dominance of the senescent phenotype in heterokaryons between replicative and post-replicative human fibroblast-like cells. Proc. Natl. Acad. Sci. USA, 71 (1974) 2231-2235. [45] R.M. Yanishevski and G.H. Stein. Ongoing DNA synthesis continues in young human diploid cells (HDC) fused to senescent HDC, but entry into S phase is inhibited. E-up. CeN Res.. 126 (1980) 465-472. [46] L.C. Drescher and J.R. Smith, Inhibition of DNA synthesis in proliferating human diploid fibroblasts by fusion with senescent cytoplasts. E.xp. Cell Res.. 144 (1983) 455-462. [47] G.C. Burmer, H. Motulsky, C.J. Zeigler and T.H. Norwood, Inhibition of DNA synthesis in young cycling human diploid fibroblast-like cells upon fusion to enucleate cytoplasts from senescent cells. E.up. Cell Res., 145 (1983) 79-84. [48] G.H. Stein and L. Atkins, Membrane-associated inhibitor of DNA synthesis in senescent human diploid fibroblasts: characterization and comparison to quiescent inhibitor. Prw. Natl. Acad. Sci. USA, 83 (1986) 9030-9034. [49] C.K. Lumpkin, J.K. McClung. O.M. Pereira-Smith and J.R. Smith, Existence of high abundance antiproliferative mRNAs in senescent human diploid fibroblasts. Science, 232 (1986) 393-395. [50] 0. Sugawara, M. Oshimura, M. Koi, L.A. Annab and J.C. Barrett. Induction of cellular senescence in immortalized cells by human chromosome 1. Science, 247 (1990) 707-710. [51] Y. Ning, J.L. Weber, A.M. Killary, D.H. Ledbetter, J.R. Smith and O.M. Pereira-Smith. Genetic analysis of indefinite division in human cells: evidence for a cell senescence-related gene(s) on human chromosome 4. Proc. Narl. Acad. Sci. USA, 88 (1991) 5635-5639.

J. Fukami et al. 1 Mechanisms of Ageing and Development 81 (1995) 139-157

157

[52] W.E. Loh Jr., H.J. Scrable, E. Livanos, M.J. Arboleda,

W.K. Cavenee, M. Oshimura and B. Weissman, Human chromosome 11 contains two different growth suppressor genes for embryonal rhabdomyosarcoma. Proc. Natl. Acud. Sci. USA, 89 (1992) 1755-1759. [53] T. Ogata, D. Ayusawa, M. Namba, E. Takahashi, M. Oshimura and M. Oishi, Chromosome 7 suppresses indefinite division of nontumorigenic immortalized human fibroblast cell line KMST-6 and SUSM-I. Mol. Cell. Biol., 13 (1993) 6036-6043. [54] A.K. Sandhu, K. Hubbard, G.P. Kaur, K.K. Jha, H.L. Ozer and R.S. Athwal, Senescence of immortal human fibroblasts by the introduction of normal human chromosome 6. Proc. Natl. Acud. Sci. USA, 91 (1994) 5498-5502. [55] J.R. Smith and O.M. Pereira-Smith,

Altered gene expression during cellular aging. Genome, 31 (1989) 386-389. [56] S. Grolstein, S. Murano, H. Benes, E.J. Moerman, A.R. Jones, R. Thweatt, R.R. Shmookler and B.H. Howard, Studies on the molecular-genetic basis of replicative senescence in Werner syndrome and normal fibroblasts. Exp. Gerontol., 24 (1989) 461-468. [57] E. Wang, I.K. Moutsatsos and T. Nakamura, Cloning and molecular characterization of a cDNA clone to statin, a protein specifically expressed in nonproliferating quiescent and senescent fibroblasts. Exp. Gerontol., 24 (1989) 485-499. [58] T. Giordano, D. Kleinsek and D.N. Foster, Increase in abundance of a transcript hybridizing to elongation factor 1 alpha during cellular senescence and quiescence. Exp. Gerontol., 24 (1989)

501--513. [59] C. Wistrom and B. Villeponteau, Cloning and expression of SAG: a novel marker of cellular senescence. Exp. Cell Res., 199 (1992) 355-362.

[60] E. Hara, T. Yamaguchi, H. Tahara, N. Tsuyama, H. Tsurui, T. Ide and K. Oda, DNA-DNA subtractive cDNA cloning using oligo(dT) 30-latex and PCR: identification of cellular genes which are overexpressed in senescent human diploid fibroblasts. Anal. Biochem., 214 (1993) 58-64. [61] V. Baldin, J. Lukas, M.J. Marcote, M. Pagan0 and G. Draetta, Cyclin D, is a nuclear protein required for cell cycle progression in G,. Genes Deu., 7 (1993) 812-821. [62] G.A. Lammie, V. Fantl, R. Smith, E. Schunring, S. Brookes, R. Michalides, C. Dickson, A. Arnold and G. Peters, D118287, a putative oncogene on chromosome I lq13, is amplified and expressed in squamous cells and mammary carcinomas and is linked to BCL-I. Oncogene, 6 (1991) 439-444. [63] W. Jiang, S.M. Kahan, N. Tomita, Y. Zhang, S. Lu and B. Weinstein, Amplification and expression of the human cyclin D gene in esophageal cancer. Cancer Res., 52 (1992) 2980-2983. [64] M. Serrano, G.J. Hannon and D. Beach, A new regulatory motif in cell cycle control causing specific inhibition of cyclin D/CDK4. Nature, 366 (1993) 704-707. [65] Y. Xiong, G.J. Hannon, H. Zhang, D. Casso, R. Kobayashi and D. Beach, p21 is a universal inhibitor of cyclin kinases. Nature, 366 (1993) 701-704. [66] K. Polyak, M.-H. Lee, H. Erdjument-Bromage, A. Koff, J.M. Roberts, P. Tempst and J. Massague, Cloning of p27”‘P’, a cyclin-dependent kinase inhibitor and a potential mediator of extracellular antimitogenic signals. Cell, 78 (1994) 59-66. [67] L. Hengst, V. Dulic, J.M. Slingerland, E. Lee and S.I. Reed, A cell cycle-regulated inhibitor of cyclin-dependent kinases. Proc. Natl. Acad Sci. USA, 91 (1994) 5291-5295. [68] A. Noda, Y. Ning, S.F. Venable, O.M. Pereira-Smith and J.R. Smith, Cloning of senescent cell derived inhibitors of DNA synthesis using an expression screen. Exp. Cell Res., 211 (1994) 90-98. [69] D.E. Quelle, R.A. Ashmun, S.A. Shurtleff, J. Kato, B. Bar-Sagi, M.F. Roussel and C.J. Sherr, Overexpression of mouse D-type cyclins accelerates G, phase in rodent fibroblasts. Genes Dev., 7 (199.3) 1559-1571. [70] M. Pagano, A.M. Theodoras, S.W. Tam and G.F. Draetta, Cyclin D,-mediated inhibition of repair and replicative DNA synthesis in human fibroblasts. Genes Dev., 8 (1994) 1627-1639.