EH

K*LD

R

ccettEfttcrlpcptiifseiby

tptHoP

2

fi4

Biochemical and Biophysical Research Communications 260, 483–487 (1999)

Article ID bbrc.1999.0883, available online at http://www.idealibrary.com on

GF-Induced Receptor Autophosphorylation in Primaryepatocytes Isolated from Phenobarbitone-Treated Mice

arin Fletcher,* Peter G. Lord,‡ Terry C. Orton,‡ J. Kevin Chipman,* and Alastair J. Strain†,1

School of Biochemistry, University of Birmingham, Edgbaston, Birmingham, B15 2TT, United Kingdom; †Liver Researchaboratories, Queen Elizabeth Hospital, Edgbaston, Birmingham, B15 2TH, United Kingdom; and ‡Safety of Medicinesepartment, Zeneca Pharmaceuticals, Alderley Park, Macclesfield, SK10 4TG, United Kingdom

eceived May 13, 1999

with a significant reduction in the responsiveness ofia4nth

(lcdnrlti

EgtbrqbraItpBhaEd

M

Cr

Phenobarbitone (PB) treatment of mice causes a de-rease in the growth factor responsiveness of hepato-ytes. Here, epidermal growth factor receptor (EGFR)xpression and receptor autophosphorylation was de-ermined in hepatocytes isolated from control and PB-reated mice. There was a decrease in the level ofGFR expression in hepatocytes isolated from mice

ollowing PB administration when compared to con-rols. EGF caused an approximate 20-fold increase ofhe 170 kD phosphotyrosine band in control hepato-ytes, which was inhibited by the EGFR specific ty-osine kinase inhibitor 4,5-dianilinopthalamide. Fol-owing PB treatment, the degree of basal receptorhosphorylation (in the absence of EGF) was signifi-antly greater and therefore the fold rise in EGFRhosphorylation in isolated hepatocytes was lowerhan in controls. However, the overall extent of EGF-nduced receptor phosphorylation was not diminishedn hepatocytes isolated from PB-treated mice. There-ore the reduction in responsiveness to growth factorseen in hepatocytes ex vivo or the cessation of prolif-ration observed in vivo following PB administrations unlikely to be attributed to a decrease in ligandinding and subsequent receptor autophosphor-lation. © 1999 Academic Press

Phenobarbitone (PB) is a non-genotoxic rodent hepa-ocarcinogen and many examples of this class of com-ound cause liver enlargement which is partially at-ributed to an initial burst of hyperplasia (1, 2).owever this effect does not persist and the cessation

f proliferation, apparent after approximately 7 daysB administration, has been shown to be associated

1 To whom correspondence should be addressed. Fax: 0121 627497. E-mail: a.j.strain@bham.ac.uk.Abbreviations used: PB, phenobarbitone; EGF, epidermal growth

actor; EGFR, epidermal growth factor receptor; TGF-a, transform-ng growth factor-a; TGF-b, transforming growth factor-b; DAPH,,5-dianilinothalamide.

483

solated hepatocytes to epidermal growth factor (EGF)nd hepatocyte growth factor (HGF) both in the rat (3,, 5) and mouse (6, 7). However, the precise mecha-isms responsible for the decreased sensitivity to mi-ogenic growth factors and the initial transient en-anced growth response remain elusive.Reductions in epidermal growth factor receptor

EGFR) numbers have been widely reported in theivers of rats administered PB, and this is often asso-iated with a decrease in ligand binding (3, 8, 9). PBoes not compete with EGF for the receptor and doesot influence receptor internalisation or EGF-inducedeceptor down regulation (10). Additionally, the centri-obular hyperplasia observed following PB administra-ion to mice is co-localised with EGFR and transform-ng growth factor-a (TGF-a) expression (11).

Autophosphorylation of tyrosine residues on theGFR is considered a critical event in EGF-mediatedrowth regulation; signal transduction via these recep-ors is directly initiated as a consequence of ligandinding and subsequent activation of cytoplasmic ty-osine kinases (12, 13). Ligand binding is most fre-uently measured using Scatchard analysis of radiola-elled ligand (3, 8, 14). However the outcome of ligandeceptor interaction can be successfully measured byssessing ligand-induced autophosphorylation (15, 16).n this investigation we have utilised a Western blot-ing method for the detection of EGF-induced phos-horylation of the EGFR in mouse hepatocytes in vitro.y evaluating the extent of receptor phosphorylation inepatocytes isolated from control and PB treated micend hence the effects of PB on this primary step of theGF signalling cascade, interpretation of the observedecrease in responsiveness to EGF may be facilitated.

ETHODS

Chemicals. Sodium phenobarbitone was obtained from Sigmahemical Company (Poole, Dorset). Collagenase A was from Boeh-inger Mannheim (Lewes, East Sussex). All media were purchased

0006-291X/99 $30.00Copyright © 1999 by Academic PressAll rights of reproduction in any form reserved.

fbu

CSdS7w

cdcwtAa4m

wieppmFdortpbAw

mtupb(teTla

R

lafauPi

alarhlbwt

t(E2

Vol. 260, No. 2, 1999 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

rom Life Technologies (Paisley, Scotland) and EGF (human recom-inant) from Sigma. All other chemicals were obtained from Sigmanless otherwise stated.

Compound administration and hepatocyte isolation. Male57Bl/6J mice were obtained from Harlan Olac Ltd (Heathfield,ussex) at approximately 6 weeks of age, housed in 12 hour light/ark cycles and allowed free access to diet (R&M, Special Dietervices). Treated mice were given PB (2000 ppm) in the diet for 3 ordays. At least four animals per group were analysed. Hepatocytesere isolated as described previously (7).

Hepatocyte treatment and sample preparation. Isolated hepato-ytes were washed twice in serum-free Williams E medium andiluted to 2 3 105 cells/ml. Aliquots of cells (tubes containing 4 3 105

ells) were either left untreated or exposed to 10 ng/ml EGF with orithout 25 mM 4,5-dianilinophalimide (DAPH) an EGFR-specific

yrosine kinase inhibitor (17), added 2 minutes prior to the EGF.fter 2 minutes incubation at 37°C, an excess of PBS (50 ml) wasdded, the cells were pelleted by centrifugation and homogenised in00 ml sodium vanadate buffer (PBS containing 10 mM EDTA, 50M NaF and 10 mM Na3VO4, pH 7.4).

Western blotting analysis. Hepatocyte homogenates were mixedith Laemmeli loading buffer, the proteins separated under reduc-

ng conditions by electrophoresis using an 8% polyacrylamide gel andlectroblotted onto a nitrocellulose membrane (Immobilon-P, Milli-ore) (9, 16). Samples prepared from control and 3 or 7 day PBre-treated mice were loaded onto the same gel and half of theembrane was subjected to immunodetection of the EGFR using the4 antibody (Sigma clone raised against the EGFR intracellularomain), followed by a peroxidase-conjugated goat anti-mouse sec-ndary antibody (Affiniti Research Products) and detected with ECLeagents (Amersham). Phosphorylated tyrosine residues were de-ected on the second half of the membrane using PT66, an anti-hosphotyrosine clone (Sigma Immunochemicals). Specific antibodyinding was detected using enhance chemiluminescence (ECL-mersham) technique as previously described (16). The membranesere stained with Ponceau stain to confirm equal protein loading.

FIG. 1. EGF receptor tyrosine phosphorylation Western blottingreatment with PB. Hepatocytes were isolated from mice by collage10 ng/ml) for 2 min. Cell homogenates were then subjected to WesteGF receptor (F4). Note EGF-induced receptor tyrosine phosphoryand 4).

484

Densitometric analysis. Lumographs were scanned and densito-etric analysis was performed using an Image Master DTS densi-

ometer and software (Pharmacia LKB). The linearity of detectionsing this method was determined previously. The ratio of the phos-horylated EGF receptor to total receptor (PT66:F4) was calculatedy dividing the densitometric value obtained from the 170 kD bandie the EGFR, as documented; 9) present on the PT66 anti-bodyreated portion of the membrane by the value obtained from thequivalent band on the F4 antibody treated portion of the membrane.his provided a semi-quantitative assessment of EGFR phosphory-

ation; a greater degree of receptor phosphorylation was indicated byn increase in the PT66:F4 value.

ESULTS

Representative data obtained with hepatocytes iso-ated from control and 3 or 7 day PB pre-treated micere illustrated in Figs. 1 and 2. The values obtainedollowing densitometric analysis of these lumographsre presented in Table 1. The mean 170 kD band val-es obtained in the absence of EGF stimulation and theT66:F4 ratios from experiments conducted are shown

n Table 2.EGF induced significant receptor phosphorylation in

ll control and PB pre-treated hepatocytes (Fig. 1,anes 2 & 4; Fig. 2, lanes 2 & 5) indicated by theppearance or augmentation of the 170 kD band rep-esenting the phosphorylated EGFR (9). This was in-ibited by the tyrosine kinase inhibitor DAPH (Fig. 2

anes 3 & 6). In the control hepatocytes the 170 kDand was more intense following EGF stimulationhen compared to the 3 or 7 day PB pre-treated hepa-

ocytes (Fig. 1 lane 2 compared with lane 4; Fig. 2 lane

epatocyte homogenates from control mice and mice following 3 daysse perfusion and were suspended in medium with or without EGFblotting using antibodies against phosphotyrosine (PT66) or againstion in hepatocytes from both control and PB-treated mice (lanes

of hnarnlat

2gT

wpc(lc(tptyt

wpc4pots

D

T

tnhEt

I(

EE(r

Vol. 260, No. 2, 1999 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

compared with lane 5) and this was confirmed by areater densitometric value of these PT66 bands (seeable 1). However, the actual amount of EGFR present

FIG. 2. EGF receptor tyrosine phosphorylation Western blottingreatment with PB. Hepatocytes were isolated from mice by collageng/ml) for 2 min. EGFr specific tyrosine kinase inhibitor DAPH (omogenates were then subjected to Western blotting using antibodGF-induced receptor tyrosine phosphorylation in hepatocytes from

yrosine phosphorylation in bith groups (lanes 3 and 6).

TABLE 1

Densitometry of Bands Present at 170 kD in Hepatocytessolated from Control, 3 Day and 7 Day PhenobarbitonePB)-Pretreated Mice

Animaltreatment

Hepatocytetreatment

PT66 bandvalue

F4 bandvalue

PT66:F4ratio

Control 0 0.1 1.8 0.1EGF 10 4.2 5.4 0.8

3 day PB 0 0.3 0.7 0.4EGF 10 1.1 0.2 5.50 0.08 2.34 0.03

Control EGF 10 4.28 4.64 0.92E10&DAPH 0.38 1.63 0.230 0.61 0.49 1.24

7 day PB EGF 10 3.29 0.73 4.48E10&DAPH 0.74 0.45 1.63

Note. Hepatocytes were left untreated (0) or treated with 10 ng/mlGF in the absence or presence of 25 mM DAPH (EGF 10,10&DAPH), subjected to Western blotting and processed with PT66

phosphorylated tyrosine) and F4 (EGFR) antibodies. Data are de-ived from the representative blot shown in Fig. 1.

485

as significantly less in the 3 day PB hepatocyte sam-les (approximately 68% reduction; Fig. 1 lanes 7–8ompared with 5–6) and further reduced after 7 days84% reduction; Fig. 2 lanes 10–12 compared withanes 7–9). Total protein loading in each lane, as indi-ated by Ponceau staining of the membranes was equalnot shown). Larger PT66:F4 values were obtained inhe presence of EGF stimulation in the 3 and 7 day PBre-treated hepatocytes (Table 2; 5.9 and 4.5 comparedo 1.2) and thus the degree of ligand-induced phosphor-lation was greater for the EGFR on these hepatocyteshan in control hepatocytes.

The basal (unstimulated) level of phosphorylationas generally greater in both the 3 and 7 day PBre-treated hepatocytes than in the control hepato-ytes. This is seen in Fig. 1 (lane 3 vs 1) and Fig. 2 (lane

vs 1). This elevation in background receptor phos-horylation was clearly indicated by the densitometryf PT66 blotts and densitometyr scan PT6:F4 ratiosaken from three experiments with cells isolated fromeparate animals (Table 2).

ISCUSSION

The role of the EGFR and its ligands (EGF andGF-a) during the PB-induced biphasic growth re-

epatocyte homogenates from control mice and mice following 7 daysperfusion and were suspended in medium with or without EGF (10

mm) was added to cells 2 min prior to the addition of EGF. Cellagainst phosphotyrosine (PT66) or against EGF receptor (F4). Noteoth control and PB-treated mice (lanes 2 and 5) and inhibition of

of hase25ies

b

s1appcttiwPl

EPshEaatcmteancPpfwE

Persistence of EGF ligand-receptor interactions ins

cElv(cTanp(pifimtf

tlptonewWtlilarai

pfllrttsctftcFgte

TABLE 2

La

C

3

7

pbmt(t

Vol. 260, No. 2, 1999 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

ponse in the liver has received much attention (3, 7, 9,1). We have used Western blotting to detect the EGFRnd to demonstrate EGF-induced receptor autophos-horylation of isolated mouse hepatocytes, an assayreviously unreported. EGF receptor specificity wasonfirmed following detection of a 170 kDa band usinghe F4 anti-EGF receptor antibody and by inhibition ofhe 170 kDa phosphorylated band by the specific EGFRnhibitor DAPH (17). Receptor phosphorylation statusas evaluated by densitometric analysis and revealedB-mediated changes in the ability of EGF to stimu-

ate autophosphorylation in hepatocytes.There was a substantial decrease in the amount ofGFR protein present in the hepatocytes isolated fromB treated mice. This is comparable with previoustudies performed using the rat; several investigatorsave demonstrated significant decreases in rat liverGFR following 10–30 day PB administration (3, 8, 9),lthough changes were not detected after 3 days PBdministration (8). The data presented here indicatehat in mouse hepatocytes there is a significant de-rease in the amount of receptor after 3 days PB ad-inistration and this has recently been confirmed in

he mouse liver using immunocytochemistry and West-rn blotting (19). The receptor down-regulation may beconsequence of ligand-binding and subsequent inter-alisation and degradation (9). The reason for the dis-repancy between the rat and mouse models after 3DB treatment remains unclear. It is possible that theeak of proliferation occurs later in the rat and there-ore an earlier onset of proliferation in the mouseould be accompanied by a more rapid reduction inGFR.

PT66:F4 Ratios, Phosphorylated Tyrosine and EGFRevels Calculated from Densitometry of Bands Presentt 170 kD

Group

PT66:F4 ratios Protein levels at 170 kD

UntreatedEGF

10 ng/mlPhosphotyrosine

PT66EGFR

F4

ontrol(n 5 5)

0.1 6 0.1 1.2 6 0.2 0.2 6 0.1 2.2 6 0.9

day PB(n 5 5)

0.6 6 0.5 5.9 6 2.9 0.3 6 0.1 0.7 6 0.4

day PB(n 5 4)

1.4 6 0.3 4.5 6 2.4 0.5 6 0.1 0.4 6 0.1

Note. Hepatocytes isolated from control, 3 day and 7 day PB-retreated mice were subjected to Western blotting following incu-ation with or without EGF (10 ng/ml). Protein levels are thoseeasured in the absence of EGF stimulation and the ratios represent

he amount of phosphorylated tyrosine (PT66) to amount of receptorF4) present. Results represent the mean 6 standard deviation fromhree experiments.

486

ub-cellular fractions has been demonstrated following125I-EGF infusion in vivo (20). In this study, hepato-ytes isolated from control mice showed negligibleGFR phosphorylation in the absence of EGF stimu-

ation, but following 3 and 7 days PB administration inivo, hepatocytes exhibited enhanced levels of basalunstimulated) phosphorylation. The paracrine/auto-rine effector of the EGFR in the liver is thought to beGF-a (21) and this is capable of binding to EGFR ondjacent cells leading to autophosphorylation and sig-al transduction (22). Elevated levels of TGF-a areresent in the mouse liver following PB administration11, 19) and it is plausible that the apparent increase inhosphorylation demonstrated herein is the outcome ofncreased receptor-ligand interactions in vivo. Thisnding supports the hypothesis that the liver enlarge-ent evident during the first week of PB administra-

ion may be in part mediated by mitogenic growthactors (19).

Investigations into the mechanisms responsible forhe reduced sensitivity of hepatocytes to mitogens fol-owing prolonged PB administration (3, 5, 6, 7) haverincipally focused on the effects of PB on EGF bindingo its receptor. Decreased EGF binding has been dem-nstrated in rats treated with PB for 2–8 weeks, witho alteration in receptor affinity (3, 10). However, oth-rs have shown an increase in EGF binding affinity,ith a concomitant decrease in receptor number (19).e have clearly demonstrated that the binding of EGF

o its receptor was not decreased in hepatocytes iso-ated mice pre-treated with PB at both time pointsnvestigated. Thus it appears that the EGFR retainigand-induced biological activity whilst hepatocytesre actively replicating (3 days PB administration;efs. 11, 19), but also when isolated hepatocytes havereduced responsiveness to mitogenic growth factors

n vitro (7 days PB treatment; ref. 7).It is feasible that the restoration of a non-

roliferative state is regulated by the inhibitory growthactor transforming growth factor-b (TGF-b), as postu-ated by Jirtle and colleagues (24). However, the lobu-ar localisation of this growth inhibitor and it’s positiveegulator, the mannose-6-phosphate receptor appearso be confined to periportal hepatocytes during thisime period (11) and therefore it appears that the ces-ation of DNA synthesis in the centrilobular hepato-ytes occurs independant of TGF-b. The observationhat the mitogenic growth factor receptors remainunctional, made in this study, suggests that perturba-ion of other elements of the receptor signalling cas-ade may be responsible for the adaptive response.urthermore, the mechanism of PB induced non-enotoxic hepatocarcinogenesis may be attributed tohe clonal expansion of cells which escape this prolif-ration control.

ACKNOWLEDGMENTS

aP

R

1

1

1

13. Sorkin, A., Helin, K., Waters, C. M., Carpenter, G., and Beguinot,

1

1

1

1

1

1

2

2

2

2

2

Vol. 260, No. 2, 1999 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

We thank Claudia Garcia-Allan and Jo Loughlin for technical helpnd acknowledge receipt of funding from the BBSRC and Zenecaharmaceuticals.

EFERENCES

1. Schulte-Hermann, R. (1979) Arch. Toxicol. Suppl. 2, 113–124.2. Greaves, P. (1996) Exp. Toxic. Pathol. 48, 169–174.3. Eckl, P. M., Meyer, S. A., Whitcombe, W. R., and Jirtle, R. J.

(1988) Carcinogenesis 9, 479–483.4. Tsai, W. H., and Michalopoulos, G. K. (1991) Cancer Lett. 57, 83–90.5. Tsai, W. H., Zarnegar, R., Michalopoulos, R., and Michalopoulos,

G. K. (1991) Cancer Lett. 59, 103–108.6. Tsai, W. H., and DeAngelo, A. B. (1996) Cancer Lett. 99, 177–183.7. Fletcher, K., Orton, T. C., Chipman, J. K., and Strain, A. J.

(1997) Arch. Toxicol. 71, 422–428.8. Gupta, C., Hatton, A., Betschart, J. M., Virji, M. A., and Shino-

zuka, H. (1988) Cancer Res. 48, 1162–1165.9. Garcia-Allan, C., Loughlin, J., Orton, T., and Lord, P. (1997)

Arch. Toxicol. 71, 409–415.0. Meyer, S. A., Gibbs, T. A., and Jirtle, R. L. (1989) Carcinogenesis

49, 5907–5912.1. Orton, T. C., Doughty, S. E., Kalinowski, A. E., Lord, P. G., and

Wadsworth, P. F. (1996) Carcinogenesis 17, 973–981.2. Ullrich, A., and Schlessinger, J. (1990) Cell 61, 203–212.

487

L. (1992) J. Biol. Chem. 267, 8672–8678.4. Boni-Schnetzler, M., and Pilch, P. F. (1987) Proc. Natl. Acad. Sci.

(USA) 84, 7832–7836.5. Zippel, R., Sturani, E., Toschi, L., Naldini, L., Alberghina, L.,

and Comoglio, P. M. (1986) Biochem. Biophys. Acta 881, 54–61.6. Gardner, M. J., Fletcher, K., Pogson, C. I., and Strain, A. J.

(1996) Biochem. Biophys. Res. Comm. 228, 238–45.7. Buchdunger, E., Trinks, U., Mett, H., Regenass, U., Muller, M.,

Meyer, T., McGlynn, E., Pinna, L. A., Trxler, P., and Lydon, N. B.(1994) Proc. Natl. Acad. Sci. (USA) 91, 2334–2338.

8. Cohen, S., Ushiro, H., Stoscheck, C., and Chinkers, M. (1982)J. Biol. Chem. 57, 1523–1531.

9. Kalinowski, A. E., Lord, P. G., Garcia-Allan, C., Jones, H. B.,Betton, G., and Orton, T. C. (1996) Human Exp. Toxicol. 15,1003.

0. Lai, W. H., Wada, C. I., Doherty, II J.-J., Kay, D. G., Posner, B. I.,and Bergeron, J. J. M. (1989) J. Cell. Biol. 109, 2741–2749.

1. Mead, J. E., and Fausto, N. (1989) Proc. Natl. Acad. Sci. (USA)86, 1558–1562.

2. Wong, S. T., Winchell, L. F., McCune, B. K., Earp, H. S., Teixido,J., Massague, J., Herman, B., and Lee, D. C. (1989) Cell 56,495–506.

3. Hwang, D. L., Ratman, A., Lev-Ran, A., and Carr, B. I. (1986)Biochem. Biophys. Res. Comms. 135, 501–506.

4. Jirtle, R. L., Hankins, G. R., Reisenbichler H., and Boyer, I. J.(1994) Carcinogenesis 15, 1473–1478.