Clinicopathologic and prognostic factors in short- and long-term surviving dogs with protein-losing enteropathy.

INTRODUCTION: The aim of the present study was to investigate the differences in the characteristics of short- and long-term surviving dogs with protein-losing enteropathy (PLE) and to identify factors that predict its outcome. We retrospectively reviewed the medical records of 59 client- owned dogs with PLE diagnosed at three different hospitals between January 2009 and November 2013. The dogs were classified as either short-term (= 6 months; STs) or long-term (> 6 months; LTs) survivors. Clinical and clinicopathological variables were investigated between the groups and receiver operating characteristic (ROC) curve analysis was performed. Nineteen dogs were classified as STs and 40 as LTs. Body weight and blood urea nitrogen concentrations were significantly higher in the STs at diagnosis (P < 0.05). At 1 month after initiation of immunosuppressive therapy (data- driven cut-off, T1), chronic canine enteropathy clinical activity index (CCECAI) scores were higher (P < 0.01) and albumin, serum total protein and total cholesterol concentrations were lower (P < 0.01) in the STs. ROC curve analysis showed that CCECAI > 5 evaluated at T1 was the best predictor of poor outcome. Although the severity of clinical signs and the majority of clinicopathological findings at diagnosis did not influence the outcome, survival time was shorter in the dogs with high CCECAI scores at T1 and which did not respond to therapy.

INTRODUCTION: Le présent travail avait pour buts d'étudier quels sont les différences de symptômes chez les chiens survivant à court et à long terme à une d'entéropathie exsudative (PLE) et d'identifier les facteurs ayant une valeur pronostique. On a étudié pour cela les dossiers médicaux de 59 chiens sur lesquels une entéropathie exsudative avait été diagnostiquée dans trois cliniques différentes entre janvier 2009 et novembre 2013. Les chiens ont été classés comme survivants à court terme (= 6 mois; STs) respectivement à long terme (= 6 mois; LTs). Les variations cliniques et clinico-pathologiques entre les groupes ont été relevées et une courbe ROC a été établie. Dixneuf chiens ont été classés comme STs et 40 comme LTs. Le poids corporel et la concentration sanguine d'urée était significativement plus élevée (P < 0.05) chez les STs que chez les LTs. Un mois après le début d'une immunosuppression (cut-off établi sur la base des données disponibles, T1), le score clinique d'activité pour une entéropathie chronique chez le chien (CCEAI) était plus élevé chez les STs que chez les LTs(P < 0.01), les valeur sanguines d'albumine, de protéines totales et de cholestérine totale par contre plus basses (P < 0.01). Dans l'analyse par la courbe ROC, un CCEAI > 5 à T1 s'est avéré être un indice fiable quant à une évolution de courte ou de longue durée. Bien que l'étendue des symptômes cliniques et la quantité des découvertes clinico-pathologiques n'aient pas influencé le pronostic, le taux de survie des chiens avec un CCEAI élevé à T1 et de ceux qui n'avaient pas répondu au traitement a été plus faible.

Constitutive overexpression of cyclin D1 does not prevent inhibition of hormone-responsive human breast cancer cell growth by antiestrogens.

Cyclin D1 is a target for positive regulation by estrogens in growth-responsive cells, in which it mediates their mitogenic effects. Amplification and overexpression of the cyclin D1 gene (CCND1) might thus represent a genetic lesion inducing hormone-independent growth of transformed cells. Indeed, cyclin D1 overexpression has been found in up to 50% of primary breast cancers, and in about one-third of these cases, this is linked to amplification of the 11q13 chromosomal region, which also includes the CCND1 gene. These tumors are predominantly estrogen receptor-positive, and for this reason, these patients are often selected for adjuvant antiestrogen therapy. No information is available, however, as to whether cyclin D1 overexpression due to gene amplification might interfere with and reduce antiestrogen efficacy. This was investigated here by taking advantage of an experimental model that reproduces cyclin D1 overexpression resulting from increased CCND1 gene dosage in hormone-responsive human breast cancer cells. For this, MCF-7 cells stably transfected with a tet-inducible cyclin D1 expression vector were tested for their in vitro response to steroidal (ICI 182,780) and nonsteroidal (trans-4-hydroxytamoxifen) antiestrogens under condition of low (endogenous only) or high (exogenous) cyclin D1 levels. Results show that although cyclin D1 overexpression seems to interfere with the early cell cycle effects of antiestrogens, it does not prevent their cytostatic actions, so that growth of cyclin-overexpressing MCF-7 cells is still efficiently inhibited in vitro by these drugs.

Estrogens do not modify MAP kinase-dependent nuclear signaling during stimulation of early G(1) progression in human breast cancer cells.

Estrogens are direct mitogens for hormone-responsive human breast cancercells, where they promote cell cycle progression and induce transcriptional activation of "immediate early" and cyclin genes. Nongenomic signaling by estrogens, including rapid changes of mitogen-activated protein(MAP) kinase and other signal-transduction-cascades activity, has been proposed to be essential for the mitogenic actions of these hormones and their nuclear receptors. Because regulation of gene transcription is considered a key step in cell cycle control by mitogenic protein kinase cascades, here we investigated the possibility that estrogen might induce the activation of extracellular signal-regulated kinase (Erk) 1/2-, c-Jun NH(2)-terminal kinase-, p38- or protein kinase A-responsive transcription factors in the cell nucleus during stimulation of early G(1) progression, a timing coincident with the maximum effects of these hormones on such enzyme activity. No significant changes in protein kinase-mediated transcription factor activity could be detected here after estrogen stimulation of either MCF-7 or ZR-75.1 cells. Furthermore, these steroids were able to induce activation of the human CCND1 gene promoter, accumulation of cyclin D1 and pRb phosphorylation, all key events in cell cycle stimulation by mitogens, even in the presence of Erk1/2 activation blockade by a MAP kinase-activating kinase (Mek)1/2 inhibitor. Thus, estrogens do not appear to convey significant protein kinase-dependent signaling to the cell nucleus during the early phases of human breast cancer cell stimulation. Furthermore, hormonal regulation of G(1) gene transcription can occur even without additional activation of the Mek-Erk1/2 pathway by estrogen receptors.

17 beta-Estradiol overcomes a G1 block induced by HMG-CoA reductase inhibitors and fosters cell cycle progression without inducing ERK-1 and -2 MAP kinases activation.

HMG-CoA reductase inhibitors, such as Lovastatin and Simvastatin, cause cell cycle arrest by interfering with the mitogenic activity of mitogens present in culture media. Cells are induced to pause in G1 and can readily resume growth upon removal of the enzymatic block. Estrogens, acting via their nuclear receptor, are mitogens for different normal and transformed cell types, where they foster cell cycle progression and cell division. In estrogen-responsive MCF-7 human breast cancer cells, but not in non responsive cells, 17 beta-estradiol (E2) induces cells arrested with Lovastatin or Simvastatin to proliferate in the presence of inhibitor, without restoring HMG-CoA reductase activity or affecting the protein prenylation pattern. Mitogenic stimulation of G1-arrested MCF-7 cells with E2 includes primary transcriptional activation of c-fos, accompanied by transient binding in vivo of the estrogen receptor and/or other factors to the ERE and the estrogen-responsive DNA region of this proto-oncogene, as detected by dimethylsulphate genomic footprinting analysis. Mitogenic stimulation of growth-arrested MCF-7 cells by E2 occurs, under these conditions, without evident activation of ERK-1 and -2 kinases, and thus independently from the mitogen-responsive signal transduction pathways that converge on these enzymes.

17beta-Estradiol induces cyclin D1 gene transcription, p36D1-p34cdk4 complex activation and p105Rb phosphorylation during mitogenic stimulation of G(1)-arrested human breast cancer cells.

MCF-7 human breast cancer cells express functional estrogen receptor and grow in response to estrogen stimulation. G(1)-synchronized MCF-7 cells, made quiescent by exposure to the HMG-CoA reductase inhibitor Simvastatin in estrogen-free medium, readily resume cell cycle progression upon stimulation with 17beta-estradiol (E(2)), even under conditions where polypeptide growth factor-triggered signal transduction pathways are inhibited by the continuous presence of Simvastatin in the culture medium. Under these conditions, cyclin D(1) gene transcription is transiently induced within the first 1-9 h of stimulation, as shown by the accumulation of cyclin D(1) mRNA and protein (p36(D(1))) in the cell and by enhanced expression of stably transfected D(1) promoter-luciferase hybrid genes. Estrogen-induced p36(D(1)) associates readily with p32(cdk2) and p34(cdk4), but not with p31(cdk5), which is however abundantly expressed in these cells. Only p36(D(1))-p34(cdk4) complexes are activated by E(2), as detected in cell extracts by immunoprecipitation with anti-D(1) antibodies followed by assessment of phosphotransferase activity toward the retinoblastoma (Rb) gene product and by analysis of p105(Rb) phosphorylation in vivo. An estrogen-responsive regulatory region has been mapped within the first 944 bp upstream of the transcriptional startsite of the human D(1) gene. Sequence analysis of this DNA region reveals that the cis-acting elements responsive to estrogen are likely to be different in this case from the canonical EREs.

Estrogen induces expression of c-fos and c-myc protooncogenes in rat uterus.

Estrogen stimulates DNA synthesis and cell proliferation in the luminal and glandular epithelia of rodent uterus. We tested the hypothesis that the mitogenic effect of estrogen occurs via activation of the expression of cellular proto-oncogenes by measuring the rate of transcription of 20 proto-oncogenes (abl, bas, erb-A, erb-B, ets, fms, fos, fps/fes, mos, myb, myc, N-myc, raf, Ha-ras, Ki-ras, N-ras, rel, sis, src, and B-lym) in the uterus of ovariectomized rats before and after injection of estrogen. c-onc transcriptional activity was monitored both by an in vitro transcription assay on isolated nuclei (run-on) and by analysis of mature mRNA. c-fos and c-myc proto-oncogenes were found to respond to estrogen with increased expression: c-fos within 30 min, with a first, sharp peak at 2 h and c-myc within 1.5 h, with a first, broad peak at 4-6 h. DNA synthesis start to increase in the uterus 13 h after estrogen injection and show a first peak at 24 h. In the liver and muscle of the same animals there is neither elevation of c-fos and c-myc expression nor increase of DNA synthesis. The kinetics of the induction by estrogen of c-fos gene expression in the uterus parallels the rate of formation of active nuclear estrogen-receptor complex. Furthermore, the ability of estrogen to induce c-fos mRNA was not abolished by the protein synthesis inhibitor cycloheximide.(ABSTRACT TRUNCATED AT 250 WORDS)

Estrogen stimulates transcription of c-jun protooncogene.

Estrogen is a mitogen for the rat uterus, where it induces transient activation of c-fos and c-myc protooncogene expression, followed by increases in DNA synthesis and cell proliferation. JUN-C, the product of the c-jun protooncogene, is a nuclear protein that can interact with FOS to modulate the activity of AP-1-responsive promoters. To test whether c-jun is a target for estrogen regulation, we measured the effects of 17 beta-estradiol on the expression of this gene in rat uterus. A human c-jun cDNA probe detects in rat uterus two mRNA species of 2.5 and 3.2 kilobases. Treatment of the animals with estrogen results in a rapid transient increase in the concentrations of these mRNAs; a 4- to 5-fold increase over the prestimulation level was detected starting 30 min after estrogen injection and lasting for 2 h, with a return to the prestimulation level after 4 h. In accordance with the results obtained by analysis of the mRNA, we found that estrogen increases 3- to 4-fold c-jun gene transcription in the uterus, at the same time it induces its mRNA accumulation. The ability of estrogen to induce c-jun gene expression was not abolished by the protein synthesis inhibitor cycloheximide, suggesting that transcriptional activation of this protooncogene is a primary response to the hormone. Furthermore, we found that in the estrogen-responsive MCF-7 human mammary carcinoma cells, estrogen stimulates transcription of a reporter gene containing four copies of a jun/AP-1 response element. These data demonstrate that c-jun gene expression is regulated by estrogen and suggest that JUN-C could play a role in the activation of cell proliferation by estrogen.

Functional antagonism between the estrogen receptor and Fos in the regulation of c-fos protooncogene transcription.

Estrogen hormones induce transient transcriptional activation of c-fos during the early phases of mitogenic stimulation of target cells. This is mediated by a functional estrogen response element (ERE) that in the human c-fos gene is localized 1kb up-stream of the transcription start site. This is the first known example of transient transcriptional activation induced by a steroid hormone acting via its nuclear receptor. Starting with the hypothesis that the product of c-fos (Fos) interferes with estrogen receptor (ER) activity on this gene promoter, generating in this way a feedback inhibition mechanism responsible for the rapid transcriptional down-regulation detected in vivo, we tested the effects of Fos overexpression on ER-mediated activation of the c-fos promoter in transfected HeLa cells. Transient transfection of an ER expression vector is followed by hormone-dependent trans-activation of reporter genes comprising the c-fos ERE linked to its own promoter. Coexpression of Fos in the cell induces a significant reduction in the activity of ER on the reporter genes. Fos antagonism is effective on both transcription activation functions of the receptor molecule and is independent of the nature of the target promoter. Furthermore, under the same experimental conditions, the estrogen-receptor complex antagonizes activation of an AP-1-responsive test gene by Fos. ER mutants deprived of the DNA-binding domain are efficient inhibitors of Fos activity, indicating that reciprocal antagonism is likely to be mediated by the formation of inactive complexes between the two factors. These results reveal the existence of a functional interference between the ER and Fos for regulation of c-fos protooncogene transcription. It is the first case in which the product of an estrogen-induced growth-related gene is shown to exert a negative feedback control on ER regulation of its own promoter.

In vivo functional analysis of the mouse estrogen receptor gene promoter: a transgenic mouse model to study tissue-specific and developmental regulation of estrogen receptor gene transcription.

Understanding the molecular and morphological basis of estrogen responsiveness in the various tissues and organs that make up an adult organism and its onset during ontogenesis requires identification of the genetic controls that determine timed expression of the estrogen receptor (ER) gene in multiple cell types. With this goal in mind, we describe here the results of the functional analysis of the mouse (m) ER gene promoter, carried out in vivo in transgenic mice. The mER gene promoter was cloned and spliced to the coding sequence of the bacterial lacZ gene (fused to the nuclear localization signal of SV40 large T: nls-beta-GAL) and then stably reintegrated into the genome of mice. Analysis of beta-GAL mRNA and protein expression in multiple organs of both female and male transgenic animals was then performed. Results show that the transgenic mER promoter, much like the endogenous one, is active in several organs and tissues of adult female and male mice. The first 0.4 kilobases of 5'-flanking DNA (up to -364) are sufficient to direct widespread expression of the transgene in mouse organs. This indicates that genetic elements functional in various cell types are included in this segment. Furthermore, the first exon and intron of the mER gene are necessary to achieve sexually dimorphic expression of the transgene in neurons located at specific sites within the central nervous system. These mER promoter transgenic mice will be useful in mapping estrogen- responsive cell types under different physiological and pathological conditions in vivo, in defining ontogenesis of estrogen action in the mouse, and in studying the mechanisms that regulate ER gene transcription.

Estrogen induces early and timed activation of cyclin-dependent kinases 4, 5, and 6 and increases cyclin messenger ribonucleic acid expression in rat uterus.

Cyclin-dependent kinases (cdks) are serine-threonine protein kinases that play a key role in the regulation of the mitotic cycle, in transcription initiation, and in the control of specific metabolic pathways in eukaryotic cells. cdk activity is controlled via phosphode-phosphorylation of the catalytic subunits of these enzymes and their physical association with cyclins and cdk inhibitors. In adult rats, estrogen stimulation results in massive proliferation of endometrial epithelial cells, accompanied by functional and structural modifications in all other tissue components of the uterus. We report here that administration of 17 beta-estradiol (E2) to adult ovariectomized rats induces within the first 25 h significant activation of cdk 4, 5, and 6, but not cdk 2, in the uterus, accompanied by increased expression of D-type (D1-3), A and E cyclin messenger RNAs (mRNAs). Furthermore, expression of the cdk inhibitor p27Kip1, a key regulator of uterine functions, is induced by E2 in this organ. Analysis of RNA extracted from E2-stimulated rat endometria shows early accumulation of D1 and D3, but not D2, cyclin mRNA, preceded by transient accumulation of c-fos mRNA. These results indicate an involvement of cdks and cyclins in estrogen actions in adult rat uterus and suggest that cyclins D1 and D3 are part of the molecular pathway that allows hormonal regulation of G1 progression in endometrial cells.

Oestrogen receptor in mammary gland cytosol of virgin, pregnant and lactating mice.

A macromolecular component binding 3H-labelled 17 beta-oestradiol in a specific manner and sedimenting in the 8-10-S region on sucrose gradient has been detected in the mammary gland cytosol of ovariectomized adult virgin mice. The dissociation constant of the macromolecule-oestradiol complex is 4.2 times 10(-10)M at 4 degrees C. The binding sites for 17beta-oestradiol of cytosol are 3.7 times 10(-14) mole/mg of protein. Incubation of cytosol with different enzymes suggests that the oestrogen-binding cytosol component is proteinaceous. The binding activity is destroyed by incubation at high temperatures and by some but not all SH-reagents tested. Competition studies show a specificity for oestrogens relative to other steroid hormones. The conclusion is that mammary gland cytosol of virgin mice contains oestradiol receptor. The receptor content does not increase in a specific manner during pregnancy and lactation but rather proportionally to total mammary gland protein.

The antiestrogen ICI 182,780 inhibits proliferation of human breast cancer cells by interfering with multiple, sequential estrogen-regulated processes required for cell cycle completion.

Antiestrogens are widely used for breast cancer treatment, where they act primarily by inhibiting the mitogenic action of estrogens on tumor cells. The effects of the pure antiestrogen ICI 182,780 on estrogen-regulated cell cycle phase-specific events were investigated here in synchronously cycling human breast cancer (HBC) cells. In early G(1)-arrested MCF-7 or ZR-75.1 cells, 17beta-estradiol (E2) induces rapid activation of the cyclin/Cdk/pRb pathway, as demonstrated by D-type G(1) cyclins accumulation during the first few hours of hormonal stimulation, followed by sequential accumulation of E, A and B1 cyclins and progressive pRb phosphorylation, as cells progress through the cell cycle. When added to quiescent cells together with E2, ICI 182,780 prevents all of the above hormonal effects. Interestingly, in mid-G(1) cells (2-8 h into estrogen stimulation) the antiestrogen causes rapid reversal of hormone-induced D-type cyclins accumulation and pRb phosphorylation, and still fully inhibits G(1)-S transition rate, while in late-G(1) cells it does not prevent S phase entry but still inhibits significantly DNA synthesis rate, S-phase cyclins accumulation and pRb hyperphosphorylation. These results indicate that pure antiestrogens prevent multiple estrogen-induced cell cycle-regulatory events, each timed to allow efficient G(1) completion, G(1)-S transition, DNA synthesis and cell cycle completion.

Transcriptional activation of jun and actin genes by estrogen during mitogenic stimulation of rat uterine cells.

Estrogens induce transcriptional activation of c-fos and c-myc proto-oncogenes during mitogenic stimulation of human, chicken, mouse and rat cells in vivo and in vitro. In this paper we show that 17 beta-estradiol injected into adult ovariectomized rats increases c-jun, jun-B and jun-D gene transcription in the uterus. Kinetics and amplitude of response are different for each gene, since c-jun is activated first, within 30 min after injection, followed by jun-D and jun-B, 60 and 90 min after injection, respectively. Maximal activation of jun-B marks a drop in transcription of all the jun genes. Furthermore, transcriptional activation by 17 beta-estradiol of the growth-regulated beta- and gamma-cytoskeletal actin genes is prevented by an inhibitor of protein synthesis, indicating that it is a secondary response to the hormone. These data support the hypothesis that during growth stimulation of target cells the estrogen receptor induces transcription of regulatory genes, triggering in this way a cascade of gene regulation events that results in progression through the cell cycle.

Estrogen regulation of proto-oncogenes coding for nuclear proteins.

Estrogen hormones are known to exert a complex influence on development and function of the female reproductive organs of vertebrates by regulating cell growth and differentiation, as well as to be implicated in oncogenesis and maintenance of tumor growth. Estrogen acts on cells via interaction with an intracellular receptor, which, like all receptors for steroid hormones, is a trans-acting transcription enhancer factor activated by the cognate ligand and capable of binding to specific, cis-acting enhancer elements usually located within the 5'-flanking regions of target genes. Additionally, estrogen regulates gene expression by influencing mRNA stability or via interaction of the estrogen receptor with transcription regulatory factors. This article reviews data indicating that estrogen directly activates (primary activation) expression of proto-oncogenes codifying for nuclear proteins that, in turn, are responsible for indirect (secondary) activation of other genes. This cascade mechanism of gene activation is likely to progress for several more steps and allows us to envisage how estrogen can direct a complex task such as cell reproduction. Among proto-oncogenes codifying for nuclear proteins, we focus on fos, jun, myc, and related genes. The mechanisms of regulation of these genes by estrogen, including regulation of transcription, messenger RNA stabilization, and protein-protein interaction, are reviewed.

Estrogen-binding proteins of calf uterus. Purification to homogeneity of receptor from cytosol by affinity chromatography.

The estrogen receptor has been purified to homogeneity from calf uterus cytosol by sequential affinity chromatography by using heparin--Sepharose 4B and 17-hemisuccinyl-17beta-estradiol-ovalbumin--Sepharose 4B. The procedure yields about 1.2 mg of receptor protein from 1 kg of calf uteri, with a recovery of 53%. The receptor protein, as a complex with 17beta-[3H]estradiol, is purified more than 99%. A single band is seen on polyacrylamide gel ectrophoresis under nondenaturing conditions. 17beta-[3H]Estradiol comigrates with the protein band. As computed from the specific activity of radioactive hormone, 64,450 g of purified receptor protein binds 1 mol of 17beta-estradiol. 17beta-[3H]Estradiol bound to the protein is displaced by estrogenic steriods but not by progesterone, testosterone, or cortisone. As judged by chromatography on calibrated Sephadex G-200 columns, the purified receptor is identical with native receptor in crude cytosol: both show a Stokes radius of 6.4 nm. On sucrose gradient in low-salt buffer, the purified receptor sediments at 8 S. On electrophoresis in NaDodSO4 gels, the purified receptor migrates as a single protein band with an apparent molecular weight of 70,000. The sedimentation coefficient measured on sucrose gradients in the presence of chaotropic salts [1 M NaBr or NaSCN (0.1 M)] is 4.2 S. We conclude that the estrogen receptor of cytosol consists of a single subunit weighing about 70,000 daltons and endowed with one estrogen binding site. Under native conditions in cytosol, several subunits associate to form a quaternary structure with a Stokes radius of 6.4 nm.

Mineralcorticoid receptor from rat kidney. Interaction with heparin and purification to a CBG-free stage.

Purification of the mineralcorticoid receptor is a particularly challenging problem. This receptor is present in target tissues at concentrations lower and is less stable than any other steroid receptor. Addition of molybdate ions (20 mM) to rat kidney cytosol enhances stability of mineralcorticoid-specific binding sites: the inactivation rate at 0 degrees C decreases from 7.2 to 1.7% per hour in the absence of aldosterone, and from 1.8 to 0.3% per hour in the presence of hormone. Rates of inactivation in the presence of molybdate are thus compatible with purification procedures. Also, the corticosteroid-binding globulin (CBG) is an important contaminating component of kidney cytosol because it cannot be specifically blocked preliminarily to affinity chromatography. We show that when kidney cytosol is incubated with heparin covalently linked to Sepharose (Sepharose-heparin), after 30 min at 0 degrees C more than 80% of the mineralcorticoid-specific binding sites interact strongly with Sepharose-heparin while CBG is not bound at all. The mineralcorticoid-specific binding sites can be recovered from Sepharose-heparin by washing with heparin (2 mg/ml; recovery up to 90%), KCl (0.3 M; recovery up to 90%); and, less efficiently, with total liver RNA (2 mg/ml; recovery up to 55%) and dextran sulfate (2 mg/ml; recovery up to 40%); little or no recovery is achieved with chondroitin sulfate, sonicated DNA, pyridoxal-5-phosphate, dextran, d-glucosamine and d-glucuronic acid. With demonstration that also the mineral-corticoid receptor binds to heparin, this property has become a general hallmark of steroid receptors. If the "heparin" binding site of steroid receptors is of physiological significance it remains to be established. By application of the newly found property of the mineralcorticoid receptor, an overall 10-fold purified, CBG-free preparation of this receptor can be obtained from kidney cytosol with a single chromatography on Sepharose-heparin.

Specific binding of estrogen receptor to sites upstream and within the transcribed region of the chicken ovalbumin gene.

By means of the DNA-cellulose competitive binding assay, the interaction of estrogen receptor complexed to 17 beta-estradiol with fragments of a cloned DNA region of the estrogen responsive chicken ovalbumin gene spanning from 1343 bps upstream to 373 bps within the transcribed region of the gene (p0V 1.7) was investigated. Only DNA fragments including either the 5'-flanking region from -21 to -140 bps or the region within the gene from +41 to +143 bps showed binding affinity for the estrogen receptor higher than calf thymus DNA. DNA fragments from human alpha 1-globin gene and glucocorticoid responsive murine mammary tumor provirus corresponding to the same DNA region investigated for ovalbumin showed affinity for the estrogen receptor no higher than that of calf thymus DNA. These results suggest that two specific binding sites for estrogen receptor are located upstream and within the ovalbumin gene, near the start-site of transcription. These receptor binding sites overlap with the 'estrogen response element' identified by Dean et al. (1) and the DNase I Hypersensitive region I found by Kaye et al. (2).