Cell-cycle-directed regulation of thymidylate synthase messenger RNA in human diploid fibroblasts stimulated to proliferate

J. Mol. Biol.

(1986) 190, 559-567

Cell-cycle-directed Regulation of Thymidylate Synthase Messenger RNA in Human Diploid Fibroblasts Stimulated to Proliferate Dai Ayusawal, Kimiko Shimizu’, Hideki Koyama’, Sumiko Kaneda’ Keiichi Takeishil and Takeshi Seno’ 1Department of Immunology and Virology Saitama Cancer Center Research Institute Ina-machi, Saitama-ken 362, Japan 2Rihara

Institute of Biological Research Yokohama City University Minami-ku, Yokohama 232, Japan

(Received 24 June 1985, and in revised form

7 April

1986)

Human diploid fibroblasts were synchronized in the resting phase by incubation in medium containing a low level of serum and then stimulated to proliferate by adding a high concentration of serum. DNA replication started 12 hours after addition of serum, and reached a maximum after 24 hours. Thymidylate synthase activity was very low in resting cells, but began to increase 12 hours after growth stimulation and thereafter continued to increase. Thymidylate synthase mRNA in the growing cells was compared with that in resting cells, using cloned human thymidylate synthase cDNA as a probe. Results showed that the mRNA content as a percentage of total RNA began to increase six hours after stimulation, reaching a level about 14 times that in unstimulated cells after 24 hours. However, the mRNA content relative to poly(A)+ RNA had increased two- to fourfold by 24 hours after growth stimulation. Transcription of the thymidylate synthase gene, determined by hybridizing labelled nascent transcripts obtained in isolated nuclei to immobilized human thymidylate synthase cDNA, was similar in the nuclei of resting and of growth-stimulated cells. These results show that the increase in thymidylate synthase mRNA in growth-stimulated cells is caused by an increase in post-transcriptional events.

1. Introduction

very low in normal liver and lymphocytes, but increases markedly when these cells are stimulated to divide, by partial hepatectomy or by exposure to phytohaemagglutinin (Maley & Maley, 1960; Haurani et d., 1978), or after neoplastic transformation (Hartmann & Heidelberger, 1861; Lockshin et al., 1979). Studies on cultured Chinese hamster cells (Conrad & Ruddle, 1972) and mouse fibroblasts (Navalgund et al., 1980) have shown that TSase activity is very low in stationary cells, but increases about 20-fold when cells are replated at low cell density or stimulated to re-enter the cell cycle by increasing the serum concentration. In mouse L1210 leukaemic cells, changes of TSase activity throughout the cell cycle have been examined in both intact cells and cell extracts (Rode et al., 1980); the TSase activity in intact cells was found to change markedly in relation to thabt of DNA synthesis, whereas the activity in cell extracts showed no change. From these and other lines of evidence, TSase is thought to be closely related to

Thymidylate synthase (EC 2.1.1.45) is essential in regulating the balanced supply of the four DNA precursors for normal DNA replication. Moreover, several lines of evidence indicate that this enzyme is involved in various genetic abnormalities, such as thymineless death (Ayusawa et al., 1983), chromosome breakages and exchanges (Hori et al., 1984a), expression of heritable fragile sites (Hori et al., 1985a), and genetic recombinations (Seno et al., 1985; Hori et al., 19846; Ayusawa et al., unpublished results), possibly by induction of a particular type of DNA double-strand break (Ayusawa et al., 1983). It has long been recognized that TSaset activity is much higher in rapidly proliferating cells than in non-cycling cells. For example, TSase activity is ‘f Abbreviations used: TSase, thymidylate synthase; kb: lo3 bases; FCS, foetal calf serum; SDS, sodium dodecyl sulphate.

559 002%2836/86/160559~09

$03.00/O

0

1986 Academic

Press Inc.

(London)

Ltd.

D. Ayusawa

560

the prolifera,tion of cells and to be the target of several widely used cancer chemotherapeutic agents. To study the genetic and biological functions of this enzyme, we have isolated various types of TSase mutants from mouse FM3A cells (Ayusawa et al., 1981a,b). Using these mutants and their derivatives, we have assigned the TSase gene (TS) Locus to human chromosome 18 (Hori et al., 1985b), determined the partial amino acid sequence of human TSase purified to homogeneity from an over-producing transformant of mouse cells (Shimizu et al.: 1985), and also isolatfed genomic DNA clones for human TSase (Takeishi et al., 1984) and functional cDNA clones (Ayusawa et al., 1984). These mutant strains and cloned DNAs now enable us to analyse the mechanism of control of expression of the TSase gene in both human and mouse cells. In this work, we examine changes in the level of human TSase mRNA and the activity of TSase during the cell cycle. To understand the mechanisms controlling the cell cycle in normal cells and in neoplastic cells, it is essential to know the early molecular events that lead resting cells to proliferate. We synchronized normal human diploid fibroblasts to the resting state (C,) in medium containing a low level of serum, and then examined changes in the level of TSase mRNA, under various conditions, following the onset of proliferation on addition of a high concentration of serum. The results suggest that expression of the TSase gene is controlled mainly by post-transcriptional events. 2. Materials

and Methods

(a) Cell lines and growth conditions Normal human fibroblasts, TIG- 1, were established from the lung of a female foetus by a project team in the Tokyo Metropolitan Institute of Gerontology (Ohashi et al., 1980). Cells at 20 population doubling levels were donated and stored frozen until used after subculture for 2 more population doublings. The cells were grown in 90 mm plastic Petri dishes (NUNCRON, Denmark) containing 10 ml of ES medium (Nissui Seiyaku Co., Ltd, Tokyo) supplemented with 100/b (v/v) FCS (GIBCO) with subculture at a split ratio of 1 : 4. The doubling time of TIG-1 cells is known to be about 20 h until a population doubling level of 60. For experiments, t,he frozen cells were thawed, cultured to subconfluency at a population doubling level of 30, washed with phosphate-buffered saline and starved of serum in ES medium supplemented with O3o/o FCS. The medium was changed on days 2 and 4; and experiments were begun on day 6 by replacing the medium by fresh ES medium supplemented with 100/b FCS to induce growth. (b) Isolation of totaE RNA and poZy(A) ’ RNA Cells were scraped off the dishes with a rubber policeman, and their t.otal RNA was isolated by the method of Chirgwin et al. (1979). Briefly, the cell pellet was lysed by vortexing in 5 ml of a lysing solution (8 Mguanidine hydrochloride, 10 mM-sodium acetate, 1 mM-

et al. dithiothreitol and 20 rnx-EDTA, adjustred to pH 5.2 with acetic acid). The resulting clear lpsate was sonicated for 30 to 60 s to shorten t~he DSA: mixed with 0.6 vol. ethanol, and stored at - 20°C. The RNA was collected by centrifugation at 15,000 g for 20 min, dissolved in 5 ml of the above lysing solution and precipitated with ethanol as described above. The RNA pellet was rinsed with ?O’j/, (v/v) cold ethanol, dried and dissolved in 2% ml of 1 rnxEDTA (pH 8.0) containing 0.19; (w/v) SDS at 65% The RNA solution was extra,cted wit,h an equal volume of chloroformjisoamyl alcohol/phenol (48 : 2 : 50; by vol.), precipitated by addition of 2 vol. ethanol after addition of 5 M-NaCl to 0.1 1\1;and kept at - 80°C for more than 2 h. Total RNA was collect,ed by centrifugation at 17,000g for 20 min, washed with cold 80% etha,nol. and stored frozen. Poly(A)+ RNA was isolated from the above tot’al RNA by chromatography on oligo(dT)-cellulose (PL Biochemicals, Inc.). The RNA sample was heated to 65°C for 5 min, mixed with an equal volume of 2 x loading buffer (40 mM-Tris . HCl (pH 7.6), I.0 If-NaCl, 2 rn$%EDTA, 0.2% SDS) and applied to a column equilibrated with the loading buffer. The eoiumn (1 cm x 1 cm) was washed extensively with the loading buffer, and t’hen RNA was eluted with 10 m&r-Tris. HCl polp(h)+ (pH 7.5), 1 rnM-EDTA, O.OSq/, SDS; mixed with sodium acetate; and precipitated with 2 vol. ethanol at -80°C. The RNA was collected by centrifugat’ion at 100,000 g for 60 min. dissolved in water. and stored frozen. (c) Northern blot hybridization The RNA sample was dissolved in an appropriate volume of a solution of 507; (v/v) formamide, 6ci/, (\-Iv) formaldehyde, 1 x MOPS buffer (20 rnx-morpholinopropane sulphonic acid, I5 rnM-sodium acetate: I mMEDTA, adjusted to pH 7.0 with acetic acid), incubated for 15 min at 6O”C, cooled on ice, and applied to I.27/, agarose gel (Seakem ME) containing 6% formaldehyde and 1 x MOPS buffer. After electrophoresis: RNA was transferred to a Gene Screen membrane (New England Nuclear) in a buffer of 25 rniw-Na,HPO,/NaW,P04 (pH 6.5). The membrane was prehybridized in prehybridization buffer (SOY/, formamide, 5 x SW @SC is 0.15 3l-KaC1, 0.015 M-sodium citrate), 50 mm ?ia,HPO,/NaH,PO, (pH 7.5), 0.1% (w/v) Ficoll. O*lOi, (w/v) bovine serum albumin; 0.1% (w/v) polyrinyipyrrolidone, 0.1% SDS, 250 ,ug sonicated and denat,ured herring sperm DNA/ml) for at least 3 h at 42°C. The membrane sheet was then hybridized to the aicktranslated TSase cDn’A probe in hybridization buffer (50% formamide, 5 x SSC. 50 mM-Na,HPO,/l NaH,PO, (pH 7.5), 0.1% Ficoll, 0.1% bovine serum albumin, O.lOi, polyvinylpyrrolidone, 0.1% SDS, 100 pg sonicated and denatured herring sperm DNA/ml, and 10% (w/v) dextran sulphate) at 42°C for more than 12 h. Then the membrane sheet was washed twice with 2 x SSC at room tempera.ture, twice with 2 x SSC containing 0.5% SDS for 30 min at 65°C and twice with 0.1 x SSC for 30 min at room temperature, and autoradiographed on an X-ray film backed with an intensifying screen at -80°C. When necessary; X-ray films were scanned with a densitometer (Bio-Rad, Model 1650) and quantified from peak areas on the chart. (d) Preparation

of

DLVA

probes

The functional cDNS clone pcHTS-1 for human TSase was isolated from an Okayama-Berg human cDNA

Expression of Thymidylate library (Ayusawa et al., 1984). Its nucleotide sequence is described elsewhere (Takeishi et al., 1985). Purified recombinant plasmid DNA was digested with restriction endonuclease BgZI and HpaI, and a resulting 1.1 kb fragment, which contained most of the coding sequence and a portion of the 3’ non-coding sequence of the cDNA insert (Fig. l(a)), was purified on agarose gel electrophoresis. DNAs were labelled by nick-translation with [a-“P]dCTP to a specific activity of 2 x 10’ cts/min per pg DNA using a Nick Translation Kit (Amersham), according to the protocol provided by the supplier,

(e) Nuclear transcription and isolation for hybridization

of

RNA

Nuclear transcription analysis was performed essentially as described by McKnight & Palmiter (1979) and Groudine et al. (1981). Cells were scraped off the dishes, collected by centrifugation, washed with ice-cold phosphate-buffered saline, and suspended in reticulocyte swelling buffer (10 mM-Tris HCl (pH 7.4) 10 mM-NaCl, 3 mM-MgCl,). Nuclei were isolated by lysis of the cells in reticulocyte swelling buffer containing 0.5% (v/v) Nonidet P-40 (Iwai Kagaku Yakuhin Co., Tokyo). The nuclei were washed several times with this lysis buffer, resuspended in a solution of 40% (v/v) glycerol, 50 mMTris. HCl (pH 8.3): 5 miv-MgCl,, 0.1 mivr-EDTA, washed once with the same solution, resuspended at a DNA concentration of 1 to 2 mg/ml in the same solution, and stored at -80°C. Reaction mixtures (200 ~1) consisted of nuclei containing about 200 pg of DNA, 30% glycerol, 2.5 mMdithiothreitol, 1 miw-MgCl,, 30 miv-KCl, 0.25 mM each of GTP and CTP, 05 mivr-ATP, and 200 /.&i of [E-‘~P]UTP (450 Ci/mmol; Amersham). The mixtures were incubated for 5 min at 26°C and then reactions were terminated by adding RNase-free DNase at 10 pg/sample (Cooper Biomedical, Inc.) and the mixtures were further incubated for 10 min at 26°C. EDTA and SDS were added to 20 mM and l%, respectively, and mixtures were incubated at 42°C with 40 pg of proteinase K (Boehringer-Mannheim) for 30 min. Then 100 ~1 of water and 100 pg of yeast tRNA were added and the mixtures were extracted twice with chloroform/isoamyl alcohol/phenol (48 : 2 : 50, by vol.) and twice with chloroform, and precipitated with ethanol. The precipitate was resuspended in 200 ~1 of 10 mivr-Tris. HCl (pH 80), 15 miw-MgCl,, 1 rnM-EDTA, and incubated for 60 min with 50 pg DNase/ml at 26°C. Final concentrations of 20 mM-EDTA and 1% SDS were added with 40 pg of proteinase K, and the mixture was incubated for 60 min at 37°C. The mixture was extracted twice with chloroform/isoamyl alcohol/phenol (48 : 2 : 50, by vol.) and twice with chloroform, and the aqueous phase was treated at 4°C with cold 10% (w/v) trichloroacetic acid in the presence of 30 mM-Na,P,G, and 1 miv-UTP to remove completely unincorporated j2P-labelled nucleotides. The precipitate was collected by centrifugation, washed 4 times with 5% trichloroacetic acid containing 10 miv-Na,P2G7, dissolved in 200 ~1 of 0.1 Msodium acetate (pH 7.0); and precipitated with ethanol. The precipitate was dissolved in 100 ~1 of 10 mMTris. HCl (pH 7.5); 5 mM-EDTA, 1% SDS, and samples were used to measure specific gene transcription by filter hybridization. The DNA probes for TSase cDNA, as described above, and for the following cDNAs or genomic DNA fragments with plasmid vector sequence were spotted onto Gene Screen, denatured with 0.2 iw-NaGH,

561

Synthase Gene

0.6 M-NaCl, washed with 0.25 M-Na,HPO,/NaH,,PO, (pH 6.5), and dried. The filters were baked for 3 h at 85”C, prehybridized in the prehybridization buffer at 42°C for 3 h; and then hybridized to labelled RNA (2 X lo6 cts/min per sample) in a final volume of 1.0 ml of the hybridization solution in a plastic vial for 72 h at 42°C with gentle shaking. The recombinant plasmids used were: pHFP-3’UT containing the 3’ untranslated region of the human /l-actin cDNA, which has no homology to other isotypic mRNAs of actin subfamilies (Ponte el! al., 1983); pH 6.3 containing human heat shock protein HSP70 gene and flanking sequences (Wu et aZ., 1985); pSVc-myc-1 containing the 2nd and 3rd exons of the cellular mouse myc gene (Land et al., 1983). The filters were washed and autoradiographed as described above. No difference was observed in the intensities of the signals obtained using DNA in B-fold and lo-fold excess of the labelled RNA (0.1 pg to 1.0 pg per filter) .with incubation times for hybridization of 48 h to 72 h. The hybridization was therefore essentially complete and saturated under these conditions of excess DNA. (f) Assay of TXase activity TSase activity was assayed radiochemically measuring tritium released from [5-3H]dUMP described by Ayusawa et al. (1981a).

by as

(g) Assay of DNA synthesis Confluent monolayers of TIG-1 cells in 35 mm plastic dishes (NUNCRON, Denmark) that had been maintained in 0.5% FCS for 6 days were stimulated to re-enter the cell cycle by addition of fresh medium containing ‘10% FCS. To stage the onset of DNA synthesis, we added 5 $Ji [3H]thymidine/ml (20 Ci/mmol; Amersham) to the medium for 30 min pulse periods at various tilmes. Incorporation of tritium into cells was terminated by removing the medium. The samples were then washed twice with ice-cold phosphate-buffered saline, twice Twith 5% trichloroacetic acid and twice with 70% ethanol, and air-dried. The materials were then solubilized with 1 ml of 1 m-NaOH/dish at room temperature and the solution was neutralized with glacial acetic acid. Samples (1 ml) were mixed with scintillation cocktail (AC13II, Amersham) and counted in a liquid scintillation counter.

3. Results (a) Characterization of human TSase mRNA in growing cells We isolated total RNA from exponentially growing cells and prepared the poly(A)+ R)NA fraction by chromatography on oligo(dT)-cellulose. To characterize human TSase mRNA in the poly(A)+ RNA fraction, we used a cloned human TSase cDNA fragment as probe. This DNA probe was derived from a BglI-HpaI fragment of a biologically active human TSase cDNA, pcHTS-1, which corresponds to the sequence from base--pair 31 to 1143 of the cDNA in which the coding sequence begins with A (numbered 1) and ends with T (numbered 939), as shown in Figure l(a) (for details see Takeishi et aE., 1985). Northern blot hybridization analysis of poly(A)+ RNA obtained from exponentially growing TIG- 1 cells revealed that a major mRNA band (96%) of 1.6 kb and two

D. Ayusawa

562

0 !

I

1

I

I

1

I

500 1

I

I

et al.

I

1

0 t

939

Bgl I I 31

I

I

b

I

!500 I

1

TAG

ATG -93

iooo 1

t

1431

HpaI I 1143

Probe

(a)

(b) I. (a) Structure of human TSase cDNA used as a probe. Relevant structural information on the func.tZom. cD?u’A clone for human TSase, pcHTS-1, is presented. The adenine residue of the translation start’ codon XTG is numbered 1. Thin lines represent t’he Okayama-Berg expression vector. In the t’hick line, the hatched region. dotted region and open region denote the coding sequence, three 5’ non-coding tandem repeats, and a 3’ untranslated region, respectively; with a poly(A)+ trail at the 3’ end. The cDKA probe is a 1113 base-pair sequence cleaved out by restriction endonucleases BgZI and HpaI. (b) Northern blot analysis of the poly(A)+ RNA fraction (5 pg) obtained from exponentially growing TIG-1 cells. The TSase cDNA fragment prepared a,s described above and in YIaterials and Methods was used as a probe.

Figure

unknown minor bands of 5.0 kb (3.5%) and 3.6 kb (0.3%) hybridized specifically to the human TSase cDNA probe (Fig. l(b)). There is indirect evidence that the 1.6 kb band is that of human TSase mRNA (Takeishi et al., 1984). The two additional bands raise the alternative possibilities of being either nuclear precursors of the cytoplasmic TSase mRNA or transcripts from a related gene. In fact, Southern blot analysis of human genomic DNA probed with the TSase cDNA fragment shows the presence of a relat’ed gene or genes (unpublished results). Although the two bands were not further characterized, it was noted that the two minor bands were faint but always present and their relative amounts remained constant during the traverse from the G, to X phase. We also used total RNA to detect human TSase mRNA and obtained a result similar to that for- t’he poly(A)+ RNA fraction. However, the hybridization signal of TSase mRNA bands was about 30-fold stronger in poly(A)+ RNA than in total RNA, when equal

amounts of RKA were processed for hybridizat,ion. This result is consistent’ with the finding, by chromatography on oligo(dT)-cellulose, that, the poly(A) * RKA content, of total RNA is about 3%, (b) Induction of TSase and DAL4 synthesis ,resting cells stimulated to proliferate

in

Various methods of obtaining cells in the resting phase (G,) were exa’mined in preliminary experiments. First, cells were grown to subeonfluency with 100/b FCS and then incubat,ed in medium containing O,l% or 0.5% serum for up to seven days. Irrespective of t’he serum concent’ration used, the rate of DSA synthesis, monitored by measuring i3H]thymidine incorporation; dropped to basal level within four to five days on serum starvation. We also determined the TSase activity in cells cultured in medium containing O*l?h or 0.5% serum for five t,o seven days, finding essentially no difference in the TSase activi’cies

Expression

of Thymidylate

Synthase

Gene

563 Time after stlmulotlon

T Q x .s < $

(h)

765-

.g g4 F 2 I & P Ii=

3-

285

2-

23s

I4 -o---

18s

I l

--0

4

8 12 16 20 Time after stimulation (h)

16s

24

Figure 2. Stimulation of resting cells to proliferate with serum. At various times after serum stimulation, cultures were pulsed for 30 min with [3H]thymidine ( [3H]dThd) and subsequent incorporation of radioactivity into DNA was determined as described in Materials and Methods. At intervals, cells were harvested, and TSase activity was assayed as described in Materials and Methods.

Time after stimulation

under either of the two conditions. Therefore, in this work we adopted the standard condition of growing cells to subconfluency with 10% serum, transferring them to medium containing 0.5% FCS for six days, and then stimulating these to proliferate by culturing them in medium containing 10% FCS. The time-course of DNA synthesis in resting cells stimulated to proliferate by this method is shown in Figure 2. DNA synthesis started about 12 hours after growth stimulation, reached a maximum rate after about 24 hours, and then decreased. This time-course was reproducible in repeated experiments, and consistent with results on WI-38 human diploid fibroblasts reported by Collins (1977). The time-course of change in TSase activity is shown in Figure 2. TSase activity in resting cells was very low, but increased steadily from about 12 hours after serum stimulation. After 24 hours, the TSase activity per cytoplasmic protein was about ten times that in resting cells. The same time-course of change of TSase activity was observed in at least four independent experiments. This increase in TSase activity was accompanied by an increase in total protein per cell, which doubled in 24 hours after growth stimulation (data not shown). (c) Human

TSase mRNA level in total RNA growth stimulation

after

Resting cells cultured in low serum for six days were stimulated to proliferate by increasing the serum concentration. Total RNA and poly(A)+ RNA were then isolated from. samples of cells harvested at intervals after growth stimulation.

0

28s

I2

24

0

(hl 12

24

-

18s --+-

(b)

Figure 3. Northern blot hybridization of TSase mRNA in resting human diploid fibroblasts stimulated to proliferate. At various times after serum stimulation, as indicated above, cells were harvested. (a) Samples of their total RNA (30 pg) were denatured, separated by electrophoresis on a 1.2% agarose gel, transferred to a membrane filter, and hybridized with excess nicktranslated human TSase cDNA probe as described in Materials and Methods. Arrows indicate the positions of mouse and Escherichia coli ribosomal RNA used as size markers. (b) Samples of their poly(A)+ RNA were isolated. Each of the left 3 lanes and the right 3 lanes contained 0.5 pg and 2.5 fig, respectively> of poly(A)+ RNA processed as in (a).

D. Ayusawa

564

et al. cells entered the growth phase, but’ its increase 24 hours after growth stimulation was two- to fourfold in several independent experiments. Since the poly(A)+ RKA content per total RNA increased TSase three- to fourfold in 24 hours, the average mRNA content per cell at this time was estimat’ed to be 20 times that in resting cells. Northern blot hybridization experimenk with both total and poly(A)+ RNA per cell, therefore, showed similar TSase mRNA levels per cell. In conclusion, after growth stimulation, the TSase mRNA level per cell increased at least one order of magnitude in 24 hours.

0

6

12

18

Time after stimulation

24 (h)

Figure 4. Plots of data obtained by scanning densitometry for the TSase mRKA contents in total RNA and poly(A)+ RNA, and nuclear transcripts in vitro. The autoradiographs shown in Figs 3, 5 and 6 were scanned with a scanning densitometer, peaks on the chart were cut out and their weights were plotted against the time of harvest of cells after growth-stimulation. (0) TSase mRn’A content of total RNA; (A) TSase mRNA content of poly (A) + R’NA; (0) run-off nuclear transcripts hybridizing with excess TSase cDNA.

Northern blot hybridization analysis of the total RNA is shown in Figure 3(a), and quant’itative data obtained by scanning densitometry are shown in Figure 4. TSase mRNA in total RNA started to increase six hours after growth stimulation and reached about 14 times that of resting cells after 24 hours. Thus; accumulation of TSase mRNA occurred before the onset of DNA replication and TSase accumulation (see Fig. 2). The RNA content per cell was found to increase gradually after growth stimulation, reaching 1.5 to 2-O times that in resting cells after 24 hours (data not shown). Therefore, the increase in TSase mRNA content per cell was 21 to 28 times that at 24 hours after stimulation. (d) TXase mRNA content of poly(Aj growth stimulation

’ R,NA after

On stimulation of resting cells, the poly(A)’ RNA content per total RNA increased, reaching three to four times the resting level in 24 hours (data not shown), whereas the total RNA content per cell increased only 1.5 to 2-O-fold in the same time. These findings are very similar to those on cultured mouse fibroblasts reported by Johnson et al. (1974). Northern blot hybridization was carried RKA fractions of cells out with the poly(A)+ harvested 0, 12 and 24 hours, respectively, after growth stimulation, as shown in Figure 3(b) (see also Fig. 4). The amount of TSase mRNA per poly(A)+ RNA fraction increased as the resting

(e) Transcription

of Tb’ase mRNA

in isolated nuclei

To determine whether the increase of TSase mRNA level on growth stimulation was due to rate or in postin transcription increase transcriptional events, we measured transcription complexes on the TSase gene in resting and growthstimulated cells. For this, nuclei were isolated from TIG-I cells at’ various stages after stimulation and incubated in the presence of [cx-~‘P]UTP and other NTPs to allow elongation of nascent RNAs present on the TSase gene. Labelled run-off transcripts were isolated and hybridized to membrane filters on which TSase cDNA was immobilized. As controis, the transcripts were hybridized to membrane filters on which recombinant plasmid DNA containing human p-a&in cDNA, human HSP70 gene, mouse o-myc gene, or bacterial aminoglycosyl-3’-phosphotransferase gene (pSV2neo) was immobilized (see ,\iIaterials and Methods). After hybridization with RNAs, the filters equal amounts of 32P-labelled were washed and processed for autoradiography. The t&al radioact’ivity incorporated into nuclei harvested from cells 24 hours after growth stimulation was t’wofold greater than that from resting cells. The result is consistent with the fact that protein and RNA content of the cell increases after serum st,imulation. In the presence of 3 pg a-amanitinlml, the total radioactivity incorporated into nuclei from cells 24 hours after serum stimulation decreased to less than 257; of t’hat in the absence of the drug (data not Bhown). Thus, most of the labelled transcripts in this reaction in vitro should have been catalysed by RKA polymerase IT. In fact, the hybridization signals in all of the following run-off assays decreased to 20yo or less when transcripts produced in the presence of 3 pg cr-amanitin/ml were used as compared to those in the absence of the drug (data not shown). The labelled nuclear transcripts hybridized very strongly to human TSase cDNA and the hybridization signals of labelled run-off transcripts to human ‘Base cDNA were a’lmost’ identical in growth-stimulated and resting cells, as shown in Figure 5(a) and Figure 4. In another experiment the hybridization signals were about twice as strong in cells 24 hours after serum stimulation as in resting cells. Thus, a,khough the amounts of

Expression of Thymidylate TSase

Synthase Gene

565

-

cDNA

(a) Time

after

0

p-

stimulation 12

(h) 24

actin

(b)

Figure 5. Hybridization of 32P-labelled RNA synthesized in nuclei in vitro to (a) human TSase cDNA and (b) other control DNAs. Nuclei were prepared from cells harvested 0, 12 and 14 h after growth-stimulation of resting cells. Nuclear transcriptions were carried out, and the 32P-labelled nuclear transcripts obtained were hybridized to the indicated DNA probes immobilized on a membrane filter as described in Materials and Methods. Samples of 1 ,ng (left) and 02 pg (right) DNA were spotted in pairs on membrane filters, to confirm that DNA was in excess of 32P-labelled run-off transcripts from the respective gene. p-actin, pHF/?A-3’UT DNA; HSP70, pH 6.3 DNA; c-myc, pSVc-myc-1 DNA; neo, pSV2neo DNA (see Materials and Methods). In (b), the DNA probes used contained plasmid DNA sequence.

cytoplasmic TSase mRNA differed at least tenfold in growth-stimulated and resting cells, the transcription rates of the TSase gene in the two conditions were within a twofold range. In contrast, the transcription rates of the B-actin gene, which has been reported to be regulated at the transcriptional level during the cell cycle (Blanchard et al., 1985), were significantly higher in cells 12 and 24 hours after serum stimulation than in resting cells (Pig. 5(b)). The hybridization signals of the transcripts to HSP70 DNA were prominent but were not different between growth-stimulated cells and resting cells in this human diploid fibroblast system. It is also noted that the labelled transcripts did not hybridize to c-myc DNA or to bacterial plasmid DNA. Therefore, the increase in the TSase mRNA level observed on growth stimulation must have been due mainly to an increase in post-transcriptional events.

4. Discussion When human diploid fibroblasts arrested in G, by a low concentration of serum were stimulated to proliferate by change to normal growth medium, protein and RNA synthesis began to increase immediately after growth stimulation, :DNA synthesis started after a lag of 12 hours, and the amounts of all three kinds of macromolecules doubled in about 24 hours for the next round of cell division. The level of mRNA, expressed on the basis of poly(A)+ RNA, increased more than that of total RNA. The poly(A)+ RNA level was twice as high in growing cells as in resting cells, even after normalization with template DNA. Under -these conditions, the TSase mRNA content relative to total RNA and poly(A)+ were, respectively, l2 to 14 times and two to four times more in growthstimulated cells than in G, cells. These different values are not contradictory, since the levlel of

566

D. Ayusawa

poly(A)+ RNA itself was three to four times higher 24 hours after growth stimulation than in resting cells. The most interesting observation in this study was that this increase in TSase mRNA, consequently resulting in induction of TSase activity, was not accompanied by an increase in the transcriptional capacity of the TSase gene in the nuclei. In fact, the transcriptional rate of the TSase gene was rather constant during progression from the resting state (GO) to the growing state (G,/S). Similar results have been reported for TSase expression in mouse cell line 3T6 (Jenh et al., 1985). Therefore, the molecular basis for the increase of TSase, one of the cell-cycle-dependent housekeeping enzymes, in growth-stimulated cells must differ from that for tissue-specific gene products, as described below. Under our condit’ions for assay of transcription in vitro in isolated nuclei: elongation and not initiation of transcription should occur, thus only nascent TSase transcripts that had been initiated in viva are measured. The hybridization signals obtained are therefore expected to be proportional to the number of RNA polymerase molecules engaged in transcription of the TSase gene. Moreover, production of the transcripts in this system was shown to be sensitive to cc-amanitin, an inhibitor of RNA polymerase II. This method has been used in many laboratories to demonstrate that expression of tissue-specific genes is regulated transcriptionally (MeKnight & Palmiter, 1979; Groudine et aZ., 1981; Powell et al., 1984). Regulation of products of growth-associated genes may occur in ways other than transcriptional control; for instance, by control of the stability of mRNA, processing of the primary transcript, or of transport of mRNA from the nucleus to the cytoplasm. Since no differences in sizes or relative amounts of the three mRNA species that hybridized to TSase cDNA were found in growing and resting cells (Figs l(b) and 3): an alternative splicing process was probably not involved in the growth-dependent regulation of TSase mRP\‘A. In the present study, we have not charact’erized the origin and nature of the two minor bands (see Fig. 2(b) and Results). It is uncertain if t’he processes of transport of mRNA from the nucleus to the cytoplasm regulate the level of TSase mRNA, as suggested by Johnson et al. (1975) for poly(A)+ RNA in grow&stimulated mouse fibroblasts. In addition to the TSase gene, growth-associated genes, for example, thymidine kinase (Groudine & Casimir, 1984), dihydrofolate reductase (Kaufman & Sharp, 1983), c-myc (Dani et al., 1985; Blanchard et aZ., 1985), histone H4 (Liischer et al., 1985) and c-myb (Thompson et al., 1986) show cell-cycledependent expression. There are large differences in the mRNA levels of these genes in dividing and non-dividing cells, but no differences in the transcript’ion rates of the genes under the two conditions, as measured by the method used here. Hence, the key step in gene expression of the above housekeeping enzymes does not seem to be

et al

transcription, as it is in the expression of tissuespecific genes as well as genes in lower euka.ryotes. At any rate, studies on the molecular mechanisms of post-transcriptional control of the TSase gene a,s well as of other related genes will be the next step in obtaining a better underst,anding of cell-cycle regulation of growth-dependent housekeeping enzymes. We thank the Cell Repository Committee of Tokyo Metropolit,an Institute of Gerontology for providing human diploid fibroblast, cells. We are indebted to Dr Mark Meuth of the Imperial Cancer Research Fund for his kind hospitality and fruitful discussion during D.A.‘s stay in his laboratory, where additional experiments for the revision of the present study were performed. We also thank Dr Larry Kedes, Dr Richard I. Morimot)o, and Dr Hartmut Land for their generous gift’s of the recombinant and pSVc-myc- 1. plasmids pHF/3A-3’UT, pH6.3. respectively. This work was supported in part by Grantsin-Aid for Cancer Research? and for Special Project, Research from the Ministry of Education, Science and Culture of Japan, and for a Comprehensive IQ-Year Strategy of Cancer Control from the Ministry of Health and Welfare of Japan.

References Ayusawa, D., Koyama;

R., Iwata, K. $ Seno, T. jl98iu).

Somat. Cell Genet. 7; 523-534.

Ayusawa; D.. Iwata. K. & Seno, T. (198lb). J. Bioi. Chem. 256; 12005-12012. Ayusawa, D., Shim& K., Koyama, II., Takeishi, K. & Seno, T. (1983). J. Biol. Chem. 258, 12448p12454. Apusawa, D., Takeishi, K., Kaneda, S.: Shimizu, K., Koyama, H. & Seno. T. (1984). J. Biol. Chem. 258. 14361-14364. Blanchard, J.-N., Piechaczyk, M.. Dani, C., Chambwd, J.-C., Franchi, A.; Pouyssegur, J. & Jeanteur. P. (1985). Xature (London), 317. 443-445. Chirgwin, J. M., Przybyla. A. E.. MacDonald. R. J. & Rutter, W. J. (1979). Biochemistry, 18, 529465299. Collins, J. M. (1977). J. Biol. Chem. 252, 141-147. Conrad, A. H. & Ruddle. F. H. (1972). J. CeZESci. 10; 471-486. Dan& Ch.. Mechti, pu’., Piechaczyk, M.. Lebleu, B.. Jeant’eur, Ph. &, Blanchard, J. M. (1985). Proc. Xa,t. Acad. Sci., U.S.A. 82, 4896-4899. Groudine, M. & Casimir, C. (1984). Nucl. Acids Res. 12, 1427-1446. Groudine, M., Peretz, M. & Weintraub. H. (1981). &‘ol. Cell. Biol. 1, 281-288. Hartmann, K.-U. 85 Heidelberger; C. (1961). J. Biol. Chem. 236, 3006-3013.

Haurani, F. I., Kardinal, C. G. & Bierma,nn; W. A. (1978). J. Cell. Physioi. 95; 49-56. Hori, T.. Apusawa, D.; Shimizu, K., Koyama, I-I. & Seno, T. (1984a). Cancer Res. 44, 703-709. Hori, T.; Ayusawa, D. & Seno. T. (19846). In S,kter Chromatid Exchanges (‘lice, R. R’. & Hollaender, A., eds). pp. 149-159, Plenum Publishing Corporation. New York. Hori, T.; Ayusawa, D., Glover, T. W. & Seno, T. (19&r*). Jpn. J. Cancer Res. 76, 977-983. Hori, T.; Ayusawa, D., Shimizu, K., Koyama, H. ej &no, T. (198%). Somat. Cell Mol. Genet. 11, 277-283. Jenh, C.-H., Geyer, P. K. & Johnson, L. F. (1985). XoE. Cell. Biol. 5, 2527-2532.

Expression of Thymidylate Johnson, L. F., Abelson, H. T.; Green, H. & Penman, S. (1974). Cell, 1, 95-100. Johnson L. F., Williams, J. G., Abelson, H. T.; Green, H. & Penman, S. (1975). Cell, 4, 69-75. Kaufman, R. J. & Sharp, P. A. (1983). Mol. Cell. Biol. 3, 1598-1608. Land, H., Parada, L. F. & Weinberg, R. A. S. (1983). Nature (London), 304, 596-602. Lockshin, A., Moran, R. G. & Danenberg, P. V. (1979). Proc. Nat. Acad. Xci., U.S.A. 76, 750-754. Liischer, B.; Stauber, C., Schindler, R. & Schiimperli, D. (1985). Proc. Nat. Acad. Sci., U.S.A. 82, 4389-4393. Maley, F. 85 Maley, G. F. (1960). J. Biol. Chem. 235, 2968-2970. &Knight, G. S. & Palmiter, R. D. (1979). J. Biol. Chem. 254, 9050-9058. Navalgund, L. G., Rossana, C., Muench, A. J. & Johnson, L. F. (1980). J. Biol. Chem. 255, 7386-7390. Ohashi, M., Aizawa, S., Ooka, H., Ohsawa, T., Kaji, K.; Kondo, H., Kobayashi, T., Nomura, T., Matsuo, M.: Mitsui, Y., Murota, S., Yamamoto, K., Ito, H., Shimada, H. & Utakoji, T. (1980). Exp. Geront. 15, 121-133.

Synthase Gene

567

Ponte, P., Gunning, P., Blau, H. & Kedes, L. (1983). MOE. Cell. Biol. 3, 1783-1791. Powell, D. J., Friedman, J. M., Oulette, A. J., Krauter, K. S. & Darnell, J. E. Jr (1984). J. Mol. Biol. 179, 21-35. Rode, W., Scanlon, K. J., Moroson, B. A. & Bertino, J. R. (1980). J. Biol. Chem. 255, 130551311. Seno, T., Ayusawa, D., Shimizu, K., Koyama, H., Takeishi, K. & Hori, T. (1985). In Genetic Consequen,ces of Nucleotide Pool Imbalance (de Serres, Publishing F. J., ed.), pp. 241-263, Plenum Corporation, New York. Shimizu, K., Ayusawa, D., Takeishi, K. & Seno, T. (1985). J. Biochem. (Tokyo), 97, 845-850. Takeishi, K., Ayusawa, D., Kaneda, S., Shimizu, K. & Seno, T. (1984). J. Biochem. (Tokyo), 95, 1477-1483. Takeishi, K., Kaneda, S., Ayusawa, D., Shimizu, K., Gotoh, 0. & Seno, T. (1985). Nucl. Acids Re;s. 13, 2035-2043. Thompson, C. B., Challoner, P. B., Neiman, P. E. & Groudine, M. (1986). Nature (London), 319, 374-380. Wu, B., Hunt, C. & Morimoto, R. (1985). Mol. Cell. Biol. 5, 330-341.

Edited by B. Mach