YY1, and the Sp Family Transcription Factors in Rabbit Endometrium Throughout Pregnancy

Archives of Medical Research 32 (2001) 263–267

ORIGINAL ARTICLE

Expression of TCF, TPF/YY1, and the Sp Family Transcription Factors in Rabbit Endometrium Throughout Pregnancy Joel Arias, Antonia Hernández, Arturo Barrón and Ivone Castro Departamento de Bioquímica y Biología Molecular, Instituto Nacional de Perinatología, Mexico City, Mexico Received for publication September 29, 2000; accepted March 26, 2001 (00/143).

Background. TCF, TPF/YY1, and the Sp family are specific transcription factors that bind sequences found within the uteroglobin (UG) gene promoter region that are necessary for transcription. To date, UG gene expression and regulation in vivo are not fully understood. The purpose of this study was to assess the expression patterns of these factors in the rabbit endometrium throughout pregnancy. Methods. Endometrial nuclear extracts were obtained from female rabbits on days 0, 3, 5, 7, 15, and 28 after mating. Transcription factor expression was assessed by DNA-protein binding assays using endometrial nuclear proteins and specific oligonucleotides. Band shifts were observed on 4% acrylamide gels and analyzed by densitometry. Results. The expression patterns of the transcription factors analyzed here differed, as TPF/YY1 and Sp3/SpR-2 were expressed constitutively while TCF and Sp1 showed variable expression patterns throughout pregnancy. Conclusions. Our results suggest that UG gene expression in the intact pregnant rabbit is controlled by two constitutive and two regulated factors, and that the DNA-binding sites are located at the TATA box and the GT1 sites within the gene promoter. © 2001 IMSS. Published by Elsevier Science Inc. Key Words: Transcription factors, Uteroglobin, Rabbit.

Introduction Uteroglobin (UG) is a small 16 kD protein. Although its physiologic function remains unknown (1), several studies have evidenced that this protein is involved in several functions such as the following: a) enhancement of blastocyst growth before implantation (2); b) progesterone transportation and protection for the developing embryo from progesterone toxicity (UG shows high affinity for this steroid) (3,4); c) immunomodulation during the implantation period in rabbits (5); d) chemotaxis and phagocytosis inhibition in monocytes and neutrophils (6,7), and e) being a phospholipase A2 (PLA2) inhibitor (8–9), it may play a role in maintaining uterus quiescence during pregnancy (10). A UG-like

Address reprint requests to: Ivone Castro, M.D., Ph.D., Departamento de Bioquímica y Biología Molecular, Instituto Nacional de Perinatología, Montes Urales 800, Col. Lomas Virreyes, 11000 México, D.F., México. Tel.: (525) 520-9900, ext. 341; FAX: (525) 520-0034; E-mail: jicastro @mail.ssa.gob.mx

secretory protein has been found in Clara cells of the lung, which also inhibits PLA2 activity and may function as a surfactant hydrolysis regulator (11). The UG gene has been cloned, sequenced, and characterized (12,13) and different steroid hormones regulate its expression in a tissue-specific manner. UG expression is regulated by progesterone in the endometrium, by estrogens in the oviduct, by glucocorticoids in the lung, and by androgens in the epididymis (14). Tissue expression is regulated by different nucleotide sequences within the gene, such as steroid response elements (SRE) and specific transcription factor binding sites located within or distant from the promoter that are essential for transcription. Different researchers have characterized these sites by footprinting techniques, electrophoretic mobility shift assay (EMSA), and by reporter gene transfections in transient expression experiments (15). Previous EMSA experiments using the OcTATA oligonucleotide (that contains the rabbit [Oryctolagus cuniculus] UG TATA box region sequence) and Ishikawa nuclear ex-

0188-4409/01 $–see front matter. Copyright © 2001 IMSS. Published by Elsevier Science Inc. PII S0188-4409(01)00 2 8 9 - 2

264

Arias et al./ Archives of Medical Research 32 (2001) 263–267

tract cultures (human endometrium adenocarcinoma) revealed the presence of two transcription factors called TCF (TATA core factor) and TPF (TATA palindrome factor) (16). The TPF sequence was determined 2 years later by Klug (1996) (17), who found that it corresponded to transcription initiation factor YY1, referred to as TPF/YY1 in this article. Sp1 and SP3/SpR-2 are other characterized transcription factors involved in activating or repressing UG gene transcription by interaction with specific GT1 (region VI of the UG promoter) sequences (18–20). The majority of transcription factor studies have been performed in vitro on mammalian and Drosophila cell lines, while very few studies have been performed in vivo in which the cells are in their natural hormonal environment. In 1996, Pérez et al. (21) demonstrated the presence of two UG gene TATA box binding factors called TRBP1 and TRBP2 in endometrial cell nuclear extracts from pseudo-pregnant rabbits. The authors suggested that in the endometrium, progesterone-induced UG gene expression is mediated by both general and tissue-specific factors. The molecular weights of these transcription factors differed from those of TPF/YY1 and TCF in Ishikawa cells. However, it was not established whether TRBPs and TPF/TCF are related proteins. Uterine fluid UG concentrations increase importantly during early pregnancy (implantation period) and decrease dramatically thereafter. There are no reports in the medical literature describing the expression patterns of these transcription factors in rabbit endometrium under normal hormonal conditions throughout pregnancy. For this reason, we assessed the expression patterns of transcription factors that bind specifically to the TATA box and GT1 sequences located within the UG gene promoter regions by EMSA, using nuclear extracts from endometrial cells sampled directly from pregnant rabbits.

Materials and Methods Preparation of Ishikawa cell nuclear extracts. A total of 80 culture plates (100-mm diameter) (Falcon, Becton Dickinson Labware, Franklin Lakes, NJ, USA) containing Ishikawa cells were left to grow to confluence in D-MEM culture medium supplemented with 10% fetal calf serum, 200 mM L-glutamine, 40 IU/mL insulin, and antibiotics. All reagents were purchased from GIBCO-BRL (Rockville, MD, USA). Cell monolayers were recovered mechanically using buffer containing 50 mM Tris (pH 7.4), 150 mM NaCl, and 1 mM EDTA (Sigma Chemical Co., St. Louis, MO, USA). Nuclei were extracted using the technique described by Klug et al. (16) with minor modifications. A protease inhibitor cocktail (0.2 mM PMSF, 0.5 mM spermidine, 0.15 mM spermine, 0.1 mM aprotinine, and 0.2 mg/m trypsin inhibitor) (GIBCO-BRL) was added to the hypotonic buffer. Aliquots were obtained from the total nuclear extract and stored at 70C until analyzed.

Extraction of nuclei from endometrial cells of pregnant rabbits. Adult New Zealand nonpregnant female rabbits weighing 3–4 kg were used for the experiment, kept under adequate temperature, light, and humidity conditions, and fed ad libitum with Purina for rabbits. The animals were mated with males previously tested for fertilization capability. Female rabbits were sacrificed in a CO2 injection chamber on day 0 (unmated rabbits) and on days 3, 5, 7, 15, and 28 after mating to extract the uterus. A total of three rabbits were sacrificed at a time. Both uterine horns were scraped to obtain endometrium, which was washed in saline solution at 4C. The cells were immediately lysed in a DOUNCE homogenizer for nucleus extraction using the technique described by Andreus et al. (22). All extracts were dialyzed with a solution containing 0.05 M HEPES, 20% glycerol, 0.5 mM EDTA, 200 mM KCl, 20 mM EGTA, and 0.5 mM DTT pH 7.6 (Sigma Chemical Co.) Nuclear extract aliquots were stored at 70C until analyzed.

Oligonucleotides. Drs. Jörg Klug and Guntram Suske of the University of Marburg, Germany kindly provided all oligonucleotides used for the present study. The sequences are described in Table 1 and are located within the UG gene promoter region (Figure 1). For oligonucleotide annealing, the mixture of both complementary sequences (10 nmol each) were heated to 90C in 100 L annealing buffer (10 mM Tris-HCl, pH 8; 1 mM EDTA, 30 mM KCl) (Sigma Chemical Co.) and left to cool at room temperature for 2 h. Double-stranded DNA was labeled with [-32P] ATP (Amersham Pharmacia Biotech, Barra Funda, São Paulo, Brasil) using the T4 polynucleotide kinase enzyme (GIBCO). Nonaligned bands were separated in a native 6% polyacrylamide (Sigma Chemical Co.) gel.

Electrophoretic mobility shift assay (EMSA). Band shift gels were prepared as described by Klug et al. (16): 5–10 g nuclear extract was pre-incubated for 10 min in ice with 100 ng unspecific poly(dI-dC) competitor, 100 mM KCl, 0.5 mM MgCl2, and 2 L 2XD (20 mM HEPES, 2 mM DTT, 0.2 mM EDTA, 17.5% glycerol) (Sigma Chemical Co.). Immediately after pre-incubation, 0.5 ng of labeled oligonucleotides (14,000 cpm) was added and the solution was incubated for an additional 15 min at room temperature. For the competitive assays, unlabeled TATA or Sp1 or random sequences were added in excess during pre-incubation on ice. Reaction mixtures (final volume  25 L) were analyzed on a 4% polyacrylamide gel and electrophoresed with specific running buffer (0.05 M Tris-base, 0.045 M boric acid, and 0.5 mM EDTA) (Sigma Chemical Co.) at 60 V for 90 min. The gel was placed on a gel drier (Model 583, BIORAD, Hercules, CA, USA) at 80C and then exposed on an X-OMAT film (Kodak, Rochester, NY, USA) for 24 h at room temperature. Shifted bands were analyzed by densitometry.

Arias et al. / Archives of Medical Research 32 (2001) 263–267 Table 1. Oligonucleotides used for EMSA Oligonucleotide ocTATA up Sp up Random

Sequence 5 AGC TTC TTG CCG GAG AAT ACA AAA AGG CAC CTC GAG G 3 5 CAC CCC TT GCC ACA CCC CTG CAC AAG 3 5 CAG CGA CTA ACA TCG ATC GC-3

265

Sp3/SpR-2 binding shows greater density and is constant throughout pregnancy. However, the density of the band produced by transcription factor Sp1 varied widely throughout pregnancy, showing a peak on day 3, decreasing on day 5, increasing once again on day 7, reaching maximum intensity on day 15, and declining again on day 28. The highest levels were observed at day 0, corresponding to unmated animals.

Results Expression patterns of transcription factors that bind to the TATA box within the UG gene promoter region. Figure 2A shows EMSA analysis of rabbit endometrium nuclear extracts in unmated rabbits (time 0) and on days 3, 5, 7, 15, and 28 postcoitus. Extracts obtained from Ishikawa cell cultures were included as controls. Two shifted bands such as those reported in Ishikawa cells are observed. Densitometric analysis (Figure 2B) shows differences in TPF/YY1 and TCF expression patterns. TPF is strongly expressed and does not vary throughout pregnancy, while TCF is weakly expressed, showing greatest intensity on days 3 and 5 and then declined (most evidently on days 7 and 15) and increased on day 28. The highest levels observed at day 28 of pregnancy are similar to day 0 (unmated animals).

Discussion We demonstrate here differences in the expression patterns of certain transcription factors that bind to the UG gene promoter TATA box, which suggest that transcription factors such as TPF/YY1 and TCF could play an important role in

Expression patterns of transcription factors binding to GT1 sequences within the UG gene promoter. Figure 3A shows that both Ishikawa and endometrial cell extracts produce the same two shifted bands, but with different expression patterns. Densitometric analysis (Figure 3B) revealed that, similar to TPF/YY1 expression, the band corresponding to

Figure 1. Uteroglobin gene promoter sequence. Underlined letters represent the sequences used for the Sp and ocTATA oligonucleotides design.

Figure 2. EMSA experiments using the UG promoter TATA box region and nuclear extracts obtained on different days of pregnancy. A) 32Plabeled TATA oligonucleotide incubated alone (oligo), incubated in the presence of Ishikawa cell nuclear extracts (Ishikawa), and incubated in the presence of endometrial cell nuclear extracts on days 0, 3, 5, 7, 15, and 28 of pregnancy. The endometrial nuclear extracts were incubated in the absence () or presence of excess (200-fold) unlabeled specific competitor (sc) and unspecific competitor (uc). B) Densitometric analysis of band shifts corresponding to reactions incubated without competitors.

266

Arias et al./ Archives of Medical Research 32 (2001) 263–267

Figure 3. EMSA experiments using the specific GT1 UG promoter sequence and nuclear extracts obtained on different days of pregnancy. A) 32 P-labeled GT1 oligonucleotide incubated alone (oligo), incubated in the presence of Ishikawa cell nuclear extracts (Ishikawa), and incubated in the presence of endometrial cell nuclear extracts on days 0, 3, 5, 7, 15, and 28 of pregnancy. The endometrial nuclear extracts were incubated in the absence () or presence of excess (200-fold) unlabeled specific competitor (sc) and unspecific competitor (uc). B) Densitometric analysis of band shifts corresponding to the reactions incubated without competitors.

UG expression throughout pregnancy in the rabbit. While TPF/YY1 was expressed constitutively, TCF expression fluctuated throughout pregnancy. We found that TCF showed highest expression levels on day 5. Because UG concentrations are highest between days 5 and 7 post-coitus (the exact time of blastocyst implantation in the rabbit) and both UG mRNA and UG concentrations remain constant during the remainder of pregnancy, several authors have suggested that UG is involved in the implantation process (23–25). The physiologic role of these transcription factors is not understood; however, because TCF is considered a transcription activator, TCF may function as a regulator of UG gene expression at critical moments, such as at blastocyst implantation. On the other hand, the fact that TPF/YY1 is constitutively expressed may be related with UG gene basal expres-

sion during the remainder of pregnancy, retaining low prostaglandin concentrations to prevent uterine contractions and premature labor. A clear negative correlation between UG and prostaglandin levels was demonstrated by Mukherjee (26), who found a sharp decline of UG concentrations and a sharp increase of prostaglandins 1 day before parturition. It was interesting to see that TCF diminished while gestation increased, at least until day 15; additionally, it increased importantly at day 28, when a low content of UG has been demonstrated. However, Peri et al. in 1995 (10) demonstrated a similar pattern expression of both UG and mRNA UG on days 18, 22, 26, and 29 of gestation in rabbits and showed an important decline at day 31. Unfortunately, they do not report results of previous days (between days 15 and 17) on which TCF diminution was observed. For that reason, we are to date unable to state clearly whether TCF is truly involved in UG expression during this period, and more research must be done in this regard. The Sp family represents another group of UG gene expression regulating transcription factors (Sp1, Sp3/SpR-2, and Sp4) that activate or inactivate the gene binding specifically to GC and CT sites within the UG promoter (19). Our results show that the Sp3/SpR-2 expression pattern is similar to that of TPF/YY1, as it was expressed constitutively. However, Sp1 expression varied throughout pregnancy, suggesting as in the case of TCF that this factor may be a transcription regulator. Several factors are known to regulate UG gene transcription in vitro. Sp1 and Sp3 are expressed at similar levels in insects, Sp1 and Sp4 act as transcription activators, and Sp3/SpR-2 functions as repressor of Sp1-mediated UG transcription in eukaryotic cell (27). Studies in endometrial rabbit tissue cultures have demonstrated the influence of some of these factors on UG gene transcription activation in vivo. Hormone-induced recruitment of Sp1 mediates estrogen activation of the rabbit UG gene in endometrial epithelium, suggesting that this factor participates in estrogen-dependent activation of this gene (28). We did not find ubiquitous Sp1 expression, which declined on pregnancy days 3 and 5. Further experiments are needed to establish whether a repressor effect of Sp3 upon Sp1 could explain this finding. Moreover, as estrogen receptors and Sp1 synergies function to induce UG gene expression after an estrogenic stimulus (28), it is possible that the levels of Sp1 (on special days of gestation) may play an important role in expression of this gene; however, this remains to be demonstrated in other experiments. On the other hand, there may be other unknown Sp1 regulating factors expressed in the intact pregnant rabbit (in contrast to pseudo-pregnancy) in which the effects of several steroid hormones or other factors are observed. It was recently described that factors TBRP1 and TBRP2 are expressed in pseudo-pregnant rabbits on day 5. Unfortunately, the authors studied TRBP1 and TRBP2 expression only on day 5 (21). We have demonstrated here that TCF and TPF/YY1 initially

Arias et al. / Archives of Medical Research 32 (2001) 263–267

described in Ishikawa cells were also found expressed in the intact pregnant rabbit on day 5. As TCF levels begin to decline after day 5, perhaps TRBP1 or TRBP2 could be involved in TCF expression regulation, but whether and how these transcription factors (TRBP1, TRBP2, TCF, and TPF/YY1) interact to regulate UG expression remain to be clarified. Our results show that the transcription factors found in Ishikawa cells are also expressed in the endometrium of the pregnant rabbit and that apparently UG gene expression was controlled throughout pregnancy by two constitutive (TPF/ YY1 and Sp3/SpR-2) and two regulated (TCF and Sp1) factors. We solely analyzed factors that bind to the TATA box and GT1 sites (region VI) localized within the UG promoter; however, other factors such as TRBP1 and TRBP2 or perhaps other yet unknown factors may be involved in regulation of this gene in vivo.

Acknowledgments This work was supported by the National Institute of Perinatology, Mexico City, Reg. #212250/07061. The authors wish to thank Drs. Jörg Klug and Guntram Suske for providing the oligonucleotides used in our experiments, as well as Susana González for her help in keeping and caring for the rabbits.

References 1. Mukherjee AB, Kundu GC, Matilde-Selvaggi G, Yuan CJ, Mandal AK, Chattopadhyay S, Sheng F, Pattabiraman N, Zhang Z. Uteroglobin: a novel cytokine? Cell Mol Life Sci 1999;55:71. 2. Krishnan RS, Daniel JC. Blastokinin inducer and regulator of blastocyst development in the rabbit uterus. Science 1967;158:490. 3. Bochskanl R, Kichner C. Uteroglobin and the accumulation of progesterone in the uterine lumen of the rabbit. Wilh Roux Arch Dev Biol 1981;190:127. 4. Bochskanl R, Thie M, Kichner C. Progesterone dependent uptake of uteroglobin by rabbit endometrium. Histochemistry 1984;80:581. 5. Mukherjee AB, Laki K, Agrawal AK. Possible mechanism of success of an allotransplantation in nature mammalian pregnancy. Med Hypothesis 1980;6:1043. 6. Shiffmann E, Geetha V, Pencev D, Warabi H, Mato J, Hirata F. Adherence and regulation of leukotaxis. Agents Actions Suppl 1983;12:106. 7. Vasanthakumar G, Manjunath R, Mukherje AB, Warabi H, Shiffmann E. Inhibition of phagocyte chemotaxis by uteroglobin, an inhibitor of blastocyst rejection. Biochem Pharmacol 1988;37:389. 8. Waite M. The phospholipases. Vol. 5. New York; Plenum;1987. 9. Dennis EA. Diversity of group types, regulation and function of phospholipase A2. J Biol Chem 1994;269:13057. 10. Peri A, Dibin NH, Dhanireddy R, Mukherjee AB. Uteroglobin gene expression in the rabbit uterus throughout gestation and in the fetal lung. Relationship between uteroglobin and eicosanoid levels in the

267

developing fetal lung. J Clin Invest 1995;96(1):343. 11. Singh G, Katial SL, Brown WE, Kennedy AL, Singh U, Wong-Chong ML. Clara cell 10 kDa protein (CC10): comparison of structure and function to uteroglobin. Biochem Biophys Acta 1990;1039:348. 12. Baily A, Atger M, Atger P, Cerbón MA, Alizon M, Vu Hai MT, Logeat F, Milgrom E. The rabbit uteroglobin gene. Structure and interaction with the progesterone receptor. J Biol Chem 1983;258(17):10384. 13. Suske G, Wenz M, Cato AC, Beato M. The uteroglobin gene region: hormonal regulation, repetitive elements and complete nucleotide sequence of the gene. Nucleic Acids Res 1983;11(8):2257. 14. Miele L, Cordella-Miele E, Mukherjee AB. Uteroglobin: structure, molecular biology, and new perspectives on its function as a phospholipase A2 inhibitor. Endocr Rev 1987;8(4):474. 15. Misseyanni A, Klung J, Suske G, Beato M. Novel upstream elements and the TATA-box region mediate preferential transcription from the uteroglobin promotor in endometrial cells. Nucleic Acids Res 1991; 19(11):2849. 16. Klug J, Knapp S, Castro I, Beato M. Two distinct factors bind to the rabbit uteroglobin TATA-box region and are required for efficient transcription. Mol Cell Biol 1994;14(9):6208. 17. Klug J, Beato M. Binding of YYI to a site overlapping a weak TATA box is essential for transcription from the uteroglobin promoter in endometrial cells. Mol Cell Biol 1996;16(11):6398. 18. Hagen G, Muller S, Beato M, Suske G. Cloning by recognition site screening of two novel GT box binding proteins: a family of Sp1 related genes. Nucleic Acids Res 1992;20(21):5519. 19. Dennig J, Hagen G, Beato M, Suske G. Members of the Sp transcription factor family control transcription from the uteroglobin promoter. J Biol Chem 1995;270(21):12737. 20. Suske G, Lorenz W, Klug J, Gazdar AF, Beato M. Elements of the rabbit uteroglobin promoter mediating its transcription in epithelial cells from the endometrium and lung. Gene Expression 1992;2(4):339. 21. Pérez MM, García C, López de Haro MS, Nieto A. Interactions of progesterone-dependent endometrial nuclear factors with the promotor of the rabbit uteroglobin gene. Arch Biochem Biophys 1996;333(1):12. 22. Andrews NC, Faller DV. A rapid micropreparation technique for extraction of DNA-binding proteins from limiting numbers of mammalian cells. Nucleic Acids Res 1991;19(9):2499. 23. Arthur AT, Cowan BD, Daniel JC. Steroid binding to blastokinin. Fertil Steril 1972;23:85. 24. Gutiérrez-Sagal R, Pérez-Palacios G, Langley E, Pasapera AM, Castro I, Cerbón MA. Endometrial expression of progesterone receptor and uteroglobin gene during early pregnancy in the rabbit. Mol Reprod Develop 1993;34(3):244. 25. Loosfelt H, Fridlansky F, Savouret JF, Atger M, Milgrom E. Mechanism of action of progesterone in the rabbit endometrium. Induction of uteroglobin and its messenger RNA. J Biol Chem 1981;256:3465. 26. Mukherjee AB, Ulane RE, Agrawal AK. Role of uteroglobin and transglutamine in masking the antigenicity of implanting rabbit embryos. Am J Reprod Immunol Microbiol 1982;2:135. 27. Hagen G, Muller S, Beato M, Suske G. Sp1-mediated transcriptional activation is repressed by Sp3. EMBO J 1994;13(16):3843. 28. Scholz A, Truss M, Beato M. Hormone-induced recruitment of Sp1 mediates estrogen activation of the rabbit uteroglobin gene in endometrial epithelium. J Biol Chem 1998;273(8):4360.