Differentiation (1984) 28 :49-52 Differentiation Q Springer-Verlag 1984

A requirement for DNA synthesis in foetal hepatocyte differentiation Effect of cytosine arabinoside on the appearance of the liver isoenzyme of pyruvate kinase

Rodney Scott and George C.T. Yeoh Department of Physiology, University of Western Australia, Nedlands, Western Australia 6009, Australia

Abstract. In hepatocyte cultures derived from 15-day-old foetal rats, the appearance of the liver (L) form of pyruvate kinase is blocked when cytosine arabinoside is added on the 2nd day of culture. When added on the 3rd day of culture, the inhibitor of DNA synthesis does not prevent the appearance of the enzyme. If cytosine arabinoside is added on the 2nd day of culture and removed on the 4th day, the enzyme is detected by the 6th day of culture. The specificity of the action of cytosine arabinoside for the L form of pyruvate kinase is in contrast with the lack of effect observed on total protein synthesis and the activity of the embryonic (M2) form of the enzyme.

Introduction

In a variety of models of cell differentiation in culture, it has been shown that the inhibition of DNA synthesis, and consequently mitosis, results in the failure of cells to differentiate and accumulate specialized proteins. For in- stance, fluorodeoxyuridine prevents the emergence of chick- embryonic red cells which synthesize haemoglobin [3]. It also blocks the appearance of definitive chondrocytes in cultures of chondrogenic cells derived from the chick limb bud [7]. Cytosine arabinoside, which also arrests DNA syn- thesis, suppresses muscle differentiation from myogenic cells derived from embryonic-chick breast muscle [12].

Foetal rat liver has been shown to differentiate in cul- ture. This is characterized by the appearance of hepatocytes which are capable of synthesizing the liver-specific enzymes, tyrosine aminotransferase [ 141 and the liver isoenzyme of pyruvate kinase (PK.L) [ll].

In a previous study [13], we have shown that foetal hepatocytes derived from 15-day-gestation rats accumulate tyrosine aminotransferase during culture. This process is almost completely abolished (95% inhibition) when cyto- sine arabinoside is added to the culture medium at the be- ginning of the experiment. In contrast, in hepatocytes de- rived from more mature foetuses (19-day gestation), which already synthesize the enzyme on day 1 of culture, cytosine arabinoside has little effect.

Holtzer et al. [4, 7-91 have proposed that DNA synthe- sis and/or the cell cycle is a prerequisite for the expression of 'new' genes which characterize the differentiated cell. In order to substantiate this, it is necessary to demonstrate that the process of blocking DNA synthesis and conse- quently arresting cells at a particular stage of the cell cycle, in this instance the GI-S interface, results in the selective suppression of the gene which would otherwise be activated.

In this study, we examined the effect of blocking DNA synthesis on the expression of two closely related genes, i.e. the embryonic (M2) and liver (L) forms of pyruvate kinase in cultured foetal rat hepatocytes. The results reveal that when cytosine arabinoside was added before the L form of the enzyme appeared, it completely abolished the appearance of the enzyme. At the same time, the M2 form, which was already present when the cultures were initiated, was unaffected. When the cultures were allowed to reach a stage at which the L form of pyruvate kinase was being synthesized, the introduction of cytosine arabinoside into the culture medium produced no effect.

Metbods Rats of the Wistar albino strain of Rattus norwegicus were used. The gestational age was determined from the time of detection of spermatazoa in the vaginal tract; this meth- od is accurate to f8 h.

'H-Thymidine and 'H-leucine were obtained from the Radiochemical Centre (Amersham, Bucks., UK). Adeno- sine 5'-diphosphate (ADP), phosphoenolpyruvate (PEP), nicotinamide adenine dinucleotide (NADH), fructose 1.6. diphosphate (FDP) and cytosine arabinoside were pur- chased from Sigma (St. Louis, MO, USA). Cell-culture me- dia, fungizone and glutamine were products of Grand Island Biological (Grand Island, NY, USA), Penicillin and streptomycin were purchased from Calbiochem (Australia). Foetal calf serum was supplied by Commonwealth Serum Laboratories. Eagle's minimal essential medium was bought from Flow Laboratories (Annandale, Australia). Coomas- sie Brilliant Blue G-250 was purchased from Bio-Rad Labo- ratories (Richmond, Calif, USA). All other reagents were of analytical grade.

Hepatocyte culture. Hepatocytes were obtained from foetal rats of 15-days gestation. The foetuses were delivered by caesarian section, and the livers were excised and placed in balanced salts solution (BSS) [51] according to the meth- od of Yeoh et al. [14]. The initial plating density of the culture derived from either 15- or 19-day-gestation liver was 2-3 x lo6 cells per 90-mm dish.

Fresh medium was given to the cultures every 24 h. Cy- tosine arabinoside was dissolved in BSS (1 mg/ml) and used at a final concentration of 7.5 pg/ml.

'H-Thymidine radioautography. On each day of culture, control and cytosine-arabinoside-treated cultures were ex- posed to 1 pCi/ml 3H-thymidine for 4 h. The cells were

50

m d

X

c ..I

then washed four times with BSS. fixed overnight with 2.5% glutaraldehyde in cacodylate buffer at 4" C and rinsed with distilled water the following day. After all of the samples had been collected, they were coated with Kodak NTB emulsion (1 : 2 v/v water) and dried in light-proof contain- ers; they were then exposed for 1 week at 4" C, developed with D19B and fixed.

Pyruvute-kinuse assay. The cells were homogenized in ho- mogenizing buffer as described previously [l 11, and the su- pernatant was used for the assay of pyruvate kinase. En- zyme activity was measured at 25" C using a coupled-assay method similar to that described by Bucher and Pfleiderer [2]. The assay buffer contained 50mM Tris (pH 7.5). 100 mM KCl, 5 mM MgSO,, 2 mM ADP, 2 mM FDP, 2 mM PEP, 0.04 mM NADH and 1 unit/ml lactate dehy- drogenase. The rate of decrease of NADH was measured at 340 nm using a Beckman DU-8 spectrophotometer. The limit of detection was found to be 0.15 nmol NADH re- duced per microgram of protein per minute.

Electrophoresis. Cellulose-acetate electrophoresis was per- formed as described previously [ll]. It was carried out for 5 h at 4" C at a constant voltage of 250 V. After electro- phoresis, the cellulose-acetate membrane was cut into 30 strips (2 mm), and the isoenzymes were eluted and assayed for activity.

Protein synthesis. Total protein synthesis was measured us- ing the method of Mans and Novelli [lo]. The cultures were exposed to 'H-leucine for 45 min and then thoroughly washed with BSS. The cells were then homogenized for 15 s using a Branson sonifier set at 1.5 A. A 5-pl aliquot of homogenate was used for the assessment of protein syn- thesis and expressed as disintegrations per minute (D.P.M.).

Protein assays were performed on all of the samples according to the method of Bearden [l] using Coomassie Brilliant Blue G-250 in a dilute acid medium. Bovine-serum- albumin (Sigma) standards were incorporated in all of the determinations.

DNA synthesis. DNA synthesis was measured by 'H-thymi- dine incorporation into DNA. Prior to harvesting, the cells were incubated for 24 h, in 'H-thymidine, washed with 5 ml of BSS (five times) and homogenized in 500 p1 trichloracetic acid 5% (TCA). The homogenate was then incubated for 1 h at 90" C and centrifuged. A 5-pl aliquot of the superna- tant was dissolved in an aqueous scintillant, and its radioac- tivity was determined in a Beckman LS 3800.

DNA assays were performed on all of the samples using the fluorimetric method described by Hinegardiner [6]; highly polymerized calf-thymus DNA (Sigma) was used as a standard.

Control

0 Cvtosine firabinoside

Results

The addition of cytosine arabinoside on day 2 of culture resulted in a marked reduction in DNA synthesis of be- tween 90% and 92% (Fig. 1). At the concentration used (7.5 pg/ml), cytosine arabinoside does not appear to be cy- totoxic, since the number of cells recovered from the cul- tures was not significantly reduced on days4, 5 and 6 in comparison with day-2 and -3 cultures (as judged by DNA recoveries from the respective cultures). Phase-contrast mi- croscope observations of the cultures also did not show

-1 1 Control

u 4 . t 0 L L

a 2 . 1 \ L

0 n

3 4 5 6 Days I n Culture

Fig. 2. The effect of cytosine arabinoside on protein synthesis. Cy- tosine arabinoside was added on day 2 of culture at a final concen- tration of 7.5 pg/ml. The incorporation of 3H-leucine into total protein was used to measure protein synthesis. Each point repre- sents the mean of eight culture dishesfSEM. The effects were tested using Student's t-test and found to be insignificant (E-O.01)

evidence of cell death. In addition, when measured under these conditions, the incorporation of 'H-leucine into the total protein in the test and control cultures was not signifi- cantly different according to the Student's t-test (Fig. 2).

In control cultures derived from either 15- or 39-day- gestation liver, 2-3 x lo6 cells attached to the culture dish after 24 h. After 4 days of culture, the number reached a maximum of about 6-7 x lo6 cells per dish. Radioauto- graphy was used to determine 'H-thymidine incorporation into hepatocytes. In this experiment, cells with hepatocyte morphology were scored as being positive or negative ac- cording to whether grains were observed over the nucleus (see Fig. 3). In controls, the percentage of positive cells was 25%, 16% and 24% on days 1, 2 and 3, respectively. In contrast, when cytosine arabinoside was added on day 2 and 'H-thymidine incorporation was assessed on day 3, no labelled cells were observed.

The effect of cytosine arabinoside on the M2 and L isoenzymes of pyruvate kinase is summarized in Fig. 4. In the control cultures, the M2 form was present at all times. The L isoenzyme could only be detected after day 3 of cul- ture. When added on day 2 of culture, the inhibitor of DNA synthesis had little effect on the level of M2 isoen- zyme. In contrast, it completely blocked the appearance of the L form of pyruvate kinase.

51

.... i 5 3 4 5 6

Daus i n Culture

Fig. 4. The effect of cytosine arabinoside on the two forms of pyru- vate kinase. Cytosine arabinoside was added on day 2 of culture at a final concentration of 7.5 pg/ml. The isoenzymes were sepa- rated electrophoretically, eluted from the cellulose-acetate mem- brane and then assayed for activity. Enzyme activity is expressed as nano moles of NADH reduced per microgram of protein per minute. Control L (A-A), test L (A-A), control M2 (H) and test M (FU). Each point represents the mean of eight culture dishes f SEM

Table 1. The effect on pyruvate-kinase activity of adding cytosine arabinoside on the 3rd day of culture

Liver form Embryonic form (L) (M2)

Fig. 3. Radioautography of cultured hepatocytes derived from 15-day-gestation rat liver using 3H-thymidine. Cultures were exposed to 'H-thymidine (1 pCi/ml) for 4 h. The cells were then washed in situ with balanced salts solution, fixed with 2.5% glutaraldehyde and prepared for radioautography. ( x 1,600)

Table 2. The effect on pyruvate-kinase activity of adding and then removing cytosine arabinoside

Liver form Embryonic form (L) (M2)

Day 3 Control ND 4.4k0.2 Cytosine arabinoside ND 4.9k0.5

Control 2.0 f 0.2 2.3 f0.4 Cytosine arabinoside 1.Of 0.2 1.5k0.3

Day 6

Cytosine arabinoside was added on day 2 of culture (final concen- tration, 7.5 pg/ml) and then removed on day 4 of culture. The enzyme activity is expressed as nano moles of NADH reduced per microgram of protein per minute. Each value represents the mean of six culture dishes f SEM. ND, not detectable

When the addition of cytosine arabinoside was delayed until day 3, the L isoenzyme appeared on day 4. The level obtained was approximately 66% of that of controls (Ta- ble 1). Table 2 shows the results of an experiment designed to test the reversibility of the suppression of L-isoenzyme appearance by cytosine arabinoside. After the mitotic inhib- itor was added on day 2, it was then removed on day 4. When the enzyme was assayed on day 6, both the M2 and L forms were present.

Day 3 Control ND 4.5 f 0.6 Cytosine arabinoside ND 3.5 f 0.4

Control 1.5k0.5 2.6k0.3 Cytosine arabinoside l.Of0.4 3.4f0.5

Day 4

Cytosine arabinoside was added at a final concentration of 7.5 pg/ ml on the 3rd day of culture. The enzyme activity is expressed as nano moles of NADH reduced per microgram of protein per minute. Each value represents the mean of eight culture dishes f SEM. ND, not detectable

Discussion The notion that DNA synthesis is necessary for cell differ- entiation [4, 7-91 has been supported by many studies em- bracing a variety of cell types. Most often, these studies have involved cells maintained in culture, since under these conditions, it is possible to block DNA synthesis without killing the cells. Arresting DNA synthesis in vivo generally leads to the death of the developing embryo. Furthermore, in culture, it is possible to remove the agent responsible for arresting DNA synthesis in order to test the reversibility

52

of the block in differentiation. Thus, it has been shown that myogenesis is dependent upon DNA synthesis [8], and the list of cell types which appear to fall in the same catego- ry include cartilage [7], red blood cells [3] and liver [13].

We have previously reported that hepatocytes derived from 15-day-gestation rat foetuses will differentiate in cul- ture with the concomitant acquisition of the liver-specific isoenzyme of pyruvate kinase [ll]. In the present study, the effect of cytosine arabinoside, an inhibitor of DNA synthesis, on this process is reported.

At the dose level selected for this study, cytosine arabin- oside is an extremely effective inhibitor of ’H-thymidine incorporation into DNA, which indicates that the synthesis of the macromolecule is largely prevented. Under these ex- perimental conditions, radioautography revealed that in control cultures, between 16% and 25% of hepatocytes in- corporated ’H-thymidine, whereas incorporation was not detectable in hepatocytes of cultures treated with cytosine arabinoside. Under these conditions, overall protein synthe- sis was essentially unaltered, yet the foetal hepatocytes failed to initiate the synthesis of the L isoenzyme of pyru- vate kinase. In these experiments, the expression of M2 embryonic isoenzyme was monitored to test further for non- specific effects of the mitotic inhibitor. This isoenzyme was detectable from the onset of culture and continued to be expressed. Altogether, these results indicate that the gene which is about to be switched on, i.e. the L-isoenzyme gene, is specifically affected.

These experiments provided no information at the cellu- lar level about the expression of PK-L and its blockade by cytosine arabinoside. It is not possible to say whether all or only a fraction of the dividing cells initiate the synthe- sis of the enzyme.

The timing of the blockade of DNA synthesis was found to be critical. When cytosine arabinoside was introduced on day 2 (2 days before the appearance of the liver-specific isoenzyme) no activity was detected on day 4. A 24-h delay in imposing the block resulted in the generation of cells which can produce the enzyme on day4. Thus, it is con- cluded the critical event [ i l l which leads to the production of cells competent to synthesize the L form of pyruvate kinase occurs between days 2 and 3 of culture. 3H-Thymi- dine-labelling experiments showed that, during this period, some hepatocytes undergo DNA synthesis, divide and sub- sequently produce the liver-specific form of pyruvate ki- nase. Previous experiments using known inducers of the pyruvate-kinase gene, such as fructose and insulin, have shown that it is not possible to alter the timing of these events by altering the environment with the aim of bringing about an early appearance of the enzyme [ll]. Therefore, it appears that the initiation of the L isoenzyme of pyruvate kinase is independent of the environment of cultured hepa- tocytes. However, i t is closely linked to DNA synthesis and/ or the mitotic history of the cells which are affected by inhibiting these processes.

Finally, these experiments demonstrate an important feature of the arrested cells, i.e. the retention of their ability to differentiate in the same manner as their control counter- parts when the DNA-synthesis arrest is removed. When hepatocytes are exposed to cytosine arabinoside and then returned to normal culture medium, the cells differentiate with respect to their ability to synthesize the L isoenzyme of pyruvate kinase. This result firstly shows that cytosine arabinoside does not selectively kill cells which would other-

wise produce the liver-specific marker, and also that hepato- cytes which recommence DNA synthesis will subsequently initiate L-isoenzyme synthesis.

In summary, the study demonstrates three important aspects of differentiation related to DNA synthesis. Firstly it is specific, because only the gene which is to be activated is involved. Secondly, the timing of events is critical, as delaying the imposition of the DNA-synthesis block results in no effect on the gene of interest. Finally, it is reversibile, because the blockade does not lead to a grossly altered cell, but one which retains its ability to produce hepatocytes capable of synthesizing the L isoenzyme of pyruvate kinase when the block is removed.

Acknowledgements. This study was supported by grants from the National Health and Medical Research Council of Australia and the Raine Foundation for the Study of Perinatal and Developmen- tal Biology.

References 1. Bearden JC Jr. (1978) Quantitation of submicrogram quantities

of proteins by an improved dye-binding assay. Biochim Biophys Acta 533 : 525-529

2. Bucher T, Pfleiderer G (1955) Pyruvate kinase from muscle. Methods Enzymol 1 :435-440

3. Campbell G LeM, Weintraub H, Holtzer H, Mayhall BH (1974) Primitive erythropoiesis in chick embryogenesis. Effect of FUdr on Hb synthesis. J Cell Physiol83 : 11-1 8

4. Dienstman SR, Holtzer H (1975) Myogenesis: A cell lineage interpretation. In: Reinert J. Holtzer H (eds) Cell cycle and differentiation, vol 7. Springer, Berlin Heidelberg New York,

5. Hanks JH, Wallace RE (1949) Relation of oxygen and tempera- ture in the preservation of tissues by refrigeration. Proc SOC Exp Biol Med 71 : 196-200

6. Hinegardner RT (1971) An improved fluorometric assay for DNA. Anal Biochem 39: 197-201

7. Holtzer H (1977) Cell lineages, stem cells and the ‘quantal’ cell cycle concept. In: Lord BI, Potten CS, Cole RJ ( 4 s ) Stem cells and tissue homeostasis. British Society for Cell Biology Symposium. Cambridge University Press, Cambridge London New York Melbourne, pp 1-28

8. Holtzer H, Weintraub H, Biehl J (1972) Cellcyclc-dependent events during myogenesis, neurogenesis and erythrogenesis. In Monroy A, Tsanov R (eds) Biochemistry of cell differentiation

9. Holtzer H, Weintraub H, Mayne R, Mochan B (1973) The cell cycle, cell lineages and cell differentiation. In Moscona A, Monroy A (eds) Current Topics in developmental biology, vol 7. Academic Press, New York, pp 229-256

10. Mans RJ, Novelli G D (1961) Measurement of the incorpora- tion of radioactive amino acids into protein by a filter-paper disk method. Arch Biochem Biophys 47:48-53

11. Scott RJ, Yeoh GCT (1983) Appearance of the liver form of pyruvate kinase in differentiating cultured foetal hepatocytes. Differentiation 28 : 64-69

12. Yeoh GCT, Holtzer H (1977) The effect of cell density, condi- tioned medium and cytosine arabinoside on myogenesis in pri- mary and secondary cultures. Exp Cell BiollO4 : 63-78

13. Yeoh GCT, Oliver IT (1980) The effect of cytosine arabinoside and bromodeoxyuridine in the appearance of tyrosine amino- transferase in cultured foetal hepatocytes of the rat. Biochem J 188:929-932

14. Yeoh GCT, Bennett FA, Oliver IT (1979) Hepatocyte differen- tiation in culture. Appearance of tyrosine aminotransfcrase. Biochem J 180:153-160

pp 1-25

24 : 41 -53

Received November 1983 / Accepted in revised form June 1984