Expression of Endogenous and Transfected Apolipoprotein II and Vitellogenin II Genes in an Estrogen Responsive Chicken Liver Cell Line

Roberta Binder, Clinton C. MacDonald, John B. E. Burch, Catherine B. Lazier, and David L. Williams Department of Pharmacological Sciences (R.B., C.C.M., D.L.W.) and Molecular Biology Graduate Program (R.B., C.C.M.) Health Sciences Center State University of New York at Stony Brook Stony Brook, New York 11794 Institute for Cancer Research (J.B.E.B.) Fox Chase Cancer Center Philadelphia, Pennsylvania 19111 Biochemistry Department (C.B.L.) Dalhousie University Halifax, Nova Scotia, B3H 4H7, Canada

A recently described chicken liver cell line, LMH, was characterized to evaluate responsiveness to estrogen. Expression of the endogenous apolipoprotein (apo) II gene was induced by 17 0-estradiol when LMH cells were cultured with chicken serum. The response was low and yielded apoll mRNA at only 0.3% of the level seen in estrogenized rooster liver. Higher levels of apoll mRNA were achieved when LMH cells were transiently transfected with an expression plasmid for estrogen receptor. A transfected apoll gene was strongly expressed only when cotransfected with receptor. Expression of the endogenous vitellogenin (VTG) II gene was not detected. However, when cotransfected with a receptor expression plasmid, VTG II reporter plasmids were expressed in LMH cells in response to 17 0estradiol. These results suggest that estrogen responsiveness of LMH cells is limited by the availability of functional receptor. Low levels of estrogen receptor mRNA were detected in LMH cells, and receptor binding sites and mRNA were greatly increased following transient transfection with a receptor expression plasmid. Using this transient transfection protocol, several VTG II reporter plasmids were compared in LMH cells and chick embryo fibroblasts. A plasmid containing VTG II estrogen response elements linked to a heterologous promoter was regulated by estrogen in both cell types. In contrast, reporter plasmids containing the VTG II promoter were regulated by estrogen in LMH cells but were not expressed at all in chick embryo fibro-

blasts. These results suggest that regulation of the VTG II gene involves cell type-specific elements in addition to estrogen response elements. LMH cells will provide a homologous cell culture model for the analysis of liver-specific and estrogen-responsive transcriptional elements in avian genes. (Molecular Endocrinology 4: 201-208, 1990)

INTRODUCTION A major difficulty in the analysis of estrogen-responsive and liver-specific genes in avian systems has been the absence of a suitable homologous cell line. Analysis of transcriptional elements in estrogen-responsive avian genes, using gene transfer procedures, have been restricted to transient assays in heterologous cell types or primary cultures (1-7). Heterologous systems and primary cultures may not be appropriate for the identification of tissue or species specific factors that regulate transcription and mRNA stability. For example, investigation of the estrogen-induced destabilization of apoll and VTG II mRNAs (8) may require longer periods of estrogen-treatment than is possible using organ or primary cell culture. Similarly, the analysis of transcription factors for estrogen-regulated and liver-specific avian genes has been problematic in that most studies have been done in heterologous systems (9-11). In order to define a causal relationship between specific factors and relevant sequences, an appropriate avian liver gene transfer system is needed. A recent report described a chicken cell line, LMH, derived from a diethylnitrosamine-induced hepatic tumor that had

0888-8809/90/0201 -0208$02.00/0 Molecular Endocrinology Copyright © 1990 by The Endocrine Society

201

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 May 2015. at 23:08 For personal use only. No other uses without permission. . All rights reserved.

Vol 4 No. 2

MOL ENDO-1990 202

some hepatocyte character (12). In the present report, the estrogen-responsiveness of this cell line has been examined. LMH cells contain a low level of functional estrogen receptor and are inducible for expression of the liver-specific apoll gene. Higher levels of expression of the endogenous apoll gene and of transfected apoll and VTG II genes were seen when LMH cells were cotransfected with an expression plasmid for estrogen receptor. Parallel transient expression studies in chick embryo fibroblast (CEF) cells provide evidence that the VTG II promoter is tissue-specific. LMH cells will provide a homologous cell culture model for the analysis of liverspecific and/or estrogen-responsive avian genes.

50 n M

EtOH L

M

l

2

3

4

1

2

3

4

1

2

3

4

RESULTS Induction of apoll mRNA by 17 /3-Estradiol We first asked whether expression of the endogenous apoll gene was responsive to estrogen in LMH cells. Apoll mRNA was measured by primer extension of an oligonucleotide complementary to nucleotides 643-664 adjacent to the poly(A) tail. As shown in Fig. 1A, apoll mRNA was detected in cells incubated with either 50 nM or 5 MM 17 /3-estradiol. Increased estrogen concentration resulted in increased apoll mRNA (compare lanes 1 grown in the presence of ethanol alone, or with 50 nM or 5 HM 17 /3-estradiol). The presence of chicken serum also was important for the magnitude of apoll mRNA induction. At either 50 nM or 5 /XM 17 /3-estradiol, induction of apoll mRNA was increased with increasing concentrations of chicken serum (compare lanes 2-4 to lanes 1, under all three conditions). The very low level of apoll mRNA seen without added hormone (lanes 1-4, EtOH) may be due to estrogen present in chicken serum or to the weak estrogenic activity of phenol red (13). By visual comparison of primer extended cDNAs using lower amounts of liver RNA to produce a signal similar to that of LMH RNA, the highest level of apoll mRNA achieved was approximately 0.3% of the maximum level seen in rooster liver (data not shown). To examine transcription initiation sites on the apoll gene, primer extension analysis was carried out with an oligonucleotide complementary to nucleotides 7090 of apoll mRNA. As shown in Fig. 1B, liver RNA (lane L) yields three initiation sites designated + 1 , —11, and +7. As previously described (14) the sites at + 1 , - 1 1 , and +7 account for 82%, 13%, and 5%, respectively, of apoll mRNA 5'-ends. The weaker bands below each start site reflect an alternate splice acceptor site in the second exon of the apoll gene (14). Apoll transcripts in LMH cells show the - 1 1 and +1 sites and the bands corresponding to the alternate splice (Fig. 1B, lanes 2 4). In the LMH cells, however, the - 1 1 site clearly predominates over the +1 site, and the +7 site was not detected. As further characterization of this cell line, we looked for other messages produced in chicken liver using northern analysis. Chicken apo A-l mRNA was detected

C 1

81

402

660

7

M Gip pp

70

90

643

664

Fig. 1. Induction of Authentic Sized apoll Transcripts A, Chicken serum improves estrogen-induced accumulation of apoll mRNA. LMH cells (2.5 x 106/100-mm dish) were incubated in Waymouth's medium containing 15% FBS (lanes 1), 1 % chicken serum (CS)/14% FBS (lanes 2), 2.5% CS/ 12.5% FBS (lanes 3), or 5% CS/10% FBS (lanes 4). After 1.5 days 17 /3-estradiol in ethanol was added to a final concentration of 50 nM or 5 ^M, as indicated. Controls contained the equivalent amount of ethanol (0.01 %). RNA was harvested 53 h after addition of estradiol and total LMH RNA (100 ^g) was primer extended with DNA oligonucleotide A (panel C). Lane L contains a primer extension of total liver RNA (4 ng) from an estrogen-induced rooster. Full-length apoll mRNA is shown by the arrowhead. Lane M contains pBR322//Wspl markers. B, Two apoll transcription start sites are identical in LMH cells and chicken liver. LMH cells (2 x 106/100 mm dish) were incubated in Waymouth's medium containing either 15% FBS (lanes 1 and 3) or 2.5% CS/12.5% FBS (lanes 2 and 4) for 1 day before addition of 17 0-estradiol (5 ^ M , lanes 3 and 4) or ethanol (lanes 1 and 2). RNA was harvested 61 h after addition of estradiol and total RNA (100 ^g) was primer extended with DNA oligonucleotide B (panel C). Lane L contains a primer extension of total RNA (4 ng) from an estrogen-induced rooster (taken from a shorter exposure time, in order to visualize the liver transcription start sites). The three transcription start sites at nucleotides - 1 1 , + 1 , and +7 are indicated. C, Oligonucleotide primers used for the primer extensions shown in panels A and B.

in a northern blot of total liver RNA and poly(A)+ LMH RNA (Fig. 2). The amount of apo A-l was relatively low compared to liver (approximately 0.2%). Chicken serum albumin mRNA was not detected by northern analysis of 20 tig total LMH RNA (data not shown). VTG II

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 May 2015. at 23:08 For personal use only. No other uses without permission. . All rights reserved.

203

An Estrogen Responsive Avian Liver Cell Line

Liver RNA

LMH RNA

transfected chicken receptor (pCEO) demonstrated that both were equally effective at inducing the transfected apoll gene (Fig. 4). Both pHEO and pCEO are in the same parent vector and are driven by the Simian virus 40 promotor. The resulting apoll mRNA transcription start sites, as well as their distribution, are the same using either receptor clone. Consequently, remaining studies used the chicken receptor clone. Measurement of Estrogen Receptor Binding Activity and mRNA

Fig. 2. Northern Analysis Detects Apolipoprotein A-l mRNA in LMH Cells Total liver RNA from an induced rooster (0.1, 0.3, and 1 ng, as labeled) and poly(A)+ RNA (selected by oligo(dT>cellulose chromatography of 20, 60, and 200 ng total LMH RNA) from LMH cells incubated in Waymouth's medium containing 5/*M 17 /3-estradiol were electrophoresed through a 1.0% Agarose/ 6% formaldehyde gel and electroblotted to a nylon membrane. The membrane was probed with a uniformly labeled single stranded apo A-l cDNA.

mRNA could not be detected by primer extension analysis of the RNA samples from Fig. 1A (data not shown). Transfection of an Estrogen Receptor Expression Plasmid Increases apoll mRNA Induction from the Endogenous Gene and a Transfected apoll Gene With an aim towards increasing induction of apoll mRNA, we employed transient transfection protocols using estrogen receptor expression vectors, a genomic apoll clone, or a combination of the two. The primer extension analysis in Fig. 3 shows that transfection of the plasmid pHEO, an expression plasmid for human estrogen receptor, increased expression of the endogenous apoll gene in LMH cells (compare lanes 1 and 2). This sample (Fig. 3, lane 2) showed that the predominant apoll transcript corresponds to the - 1 1 transcription start site as was seen without transfection of pHEO (Fig. 1B). Transfection of only the apoll genomic clone pApo107 did not increase induction over that seen in the control transfection (compare lanes 1 and 3 in the [+] estrogen panels). Transfection of both pHEO and pApo107 dramatically increased the amount of apoll induction over the that seen by transfection of receptor alone (compare lanes 2 and 4 in the [+] estrogen panels). As opposed to the endogenous apoll mRNAs, the transfected apoll mRNAs appear to have all three transcription start sites that were mapped in liver apoll mRNA. In addition, the distribution of transcription start sites is similar to that seen in liver in which the +1 site predominates (compare Fig. 3, lanes 4 to Fig. 1B, lane L). Unlike the endogenous receptor, the transfected receptor induced both the endogenous and transfected apoll genes equally well with either 50 nM or 5 HM estradiol. A comparison of transfected human receptor and

The weak estrogenic response of the endogenous apoll gene and the dependence of the transfected apoll gene upon cotransfection with a receptor expression plasmid suggested that LMH cells are limited in the content of active receptor. To verify that active estrogen receptor was increased after transfection of the receptor plasmid, both ligand binding activity and receptor mRNA levels were examined in nontransfected LMH cells and in cells transfected with pCEO. Transfected cell also received pApo107 in order to confirm apoll induction in this experiment (data not shown). Cells transfected with pCEO contained 260 fmol/mg protein and 36 fmol/mg protein of receptor binding sites in nuclear and cytosolic extracts, respectively. Scatchard analysis of the cytosolic extract indicated a dissociation constant (Kd) = 2 4 nM (data not shown). With the amounts of extract assayed, receptor activity was not detected in nontransfected LMH cells (