Modulation of gap junction transcript and protein expression during pregnancy in the rat

The expression of three different gap junction transcripts, or, (Cx43), B~ (Cx39, and B2 (Cx26) was examined in several organs during pregnancy in the rat. In all of the organs that were examined-uterus, ovary, heart, and l iverthere was a strong correlation between levels of gap junction mRNA and gap junction antigens that were detected at different stages of pregnancy. A striking change in o~ transcript levels (a 5.5-fold increase) was detected in the uterine myometrium on the day before parturition. This elevation of the c~, transcript is thought to be associated with the formation of gap junctions that are required for synchronizing the contractility of the myometrial cells during parturition. 2 d before parturition, there was a detectable elevation of/32 transcripts and protein in the endometrial epithelium, which was then followed by a dramatic decrease in j52 gap junctional protein on the day before parturition. There was also a substantial elevation of tx~ transcripts (a 6.7-fold increase) in the stromal regions of the ovary on the day before parturition that was identical to the temporal pattern of ott expression in the myometrium. In all three instances-the c~, transcripts in the myometrium, ~2 transcripts in the endometrium, and at transcripts in the ovarythe transcript modulation appeared to be cell specific, because the changes in transcript levels of these three gene products occurred independently of the poly(A)+RNA concentrations at the same pregnancy stages in the respective organs. There were no specific changes detected in gap junction transcript levels in the heart and liver during pregnancy. These observations indicate that a cell-specific modulation of gap junction expression occurs in two regions of the uterus and the ovary during pregnancy. Further, it appears that the same gap junction gene in different organs, such as the ct~ gene in the uterine myometrium and the heart, can be differentially regulated. I T is now well established that cell-to-cell communication in various tissues is mediated by specialized membrane regions known as gap junctions (Gilula et al., 1972; Loewenstein, 1981). Recent molecular characterizations have provided direct evidence for a family of related gap junction (GJ) t proteins that form intercellular channels. This includes the identification of cDNA clones coding for three different GJ proteins with deduced molecular masses of 32 kD from mammalian liver (Paul, 1986; Kumar and Gilula, 1986), 43 kD from mammalian heart (Beyer et al., 1987), and a 26-kD protein from mammalian liver (Nicholson and Zhang, 1988). A fourth candidate is a 70-kD junction protein component of the lens fiber junctions that has an amino terminus that is homologous to those of the liver and heart GJ proteins (Kistler et al., 1988; Beyer et al., 1989). Because all of these coding sequences have a conserved region of"o 200 amino-terminal amino acid residues containAddress correspondence to Dr. Norton B. Gilula, Department of Molecular Biology, Research Institute of Scripps Clinic, 10666 N. Torrey Pines Road, La Jolla, CA 92037. 1. Abbreviations used in this paper: G J, gap junction; KLH, keyhole limpet hemocyanin. ing four putative transmembrane stretches with the aminoand carboxy-terminal residues likely to be located on the cytoplasmic side of the junction membrane, it has been possible to deduce a generalized structure for junctional membrane protein topology (Zimmer et al., 1987; Beyer et al., 1987; Milks et al., 1988; Goodenough et al., 1988; Nicholson and Zhang, 1988; Hertzberg et al., 1988; Yancey et al., t989). It has been known for some time that there are significant changes in GJ populations in various female reproductive organs during hormone-dependent events. For example, the mammalian preovulatory ovarian follicle contains a large population of GJs that connect granulosa cells to each other, as well as to the oocyte itself (Gilula et al., 1978; Larsen et al., 1981). The ovarian follicle undergoes dramatic changes in structure and activity during the menstrual cycle and pregnancy. An apparent downregulation of GJs has been reported during early and late preovulatory periods for the follicular cells and the cumulus-oophorus in response to an ovulatory stimulus (Larsen et al., 1981; Larsen et al., 1986), indicating that GJ structures can be regulated differentially within the same follicular unit. © The Rockefeller University Press, 0021-9525190/021269/14 $2.00 The Journal of Cell Biology, Volume 110, February 1990 269-282 269 on July 6, 2017 jcb.rress.org D ow nladed fom GJs have been also documented to provide low-resistance pathways for electrical coupling between neighboring cells in several smooth muscle cell systems (for review see Daniel, 1987). This includes the uterine myometrium, which has been shown to dramatically increase its GJ content at parturition (term) (Garfield et al., 1977, 1978, 1982; Dahl and Berger, 1978; Saito et al., 1985; Ikeda et al., 1987) or after estrogen treatment of immature animals (Burghardt et al., 1984; Garfield et al., 1978). In addition to these morphological observations, a GJ transcript has been detected in term uterus (Beyer et al., 1987) and a major 43-kD protein has been detected in uterine samples using peptide antibodies prepared to a region of the rat heart GJ protein (Dupont et al, 1988; Beyer et al., 1989). The present system that has been used to distinguish the different GJ gene products is based principally on the molecular weights of the deduced protein products (Beyer et al., 1987). Unfortunately, this approach has limitations because protein products of many different sizes have been reported (Table I), and similar genes in different organisms can produce protein products of slightly different sizes; i.e., the mammalian liver gene product is 32 kD (Paul, 1986; Kumar and Gilula, 1986), whereas the Xenopus liver gene product is 30 kD (Gimlich et al., 1988). Consequently, we have applied a Greek nomenclature that emphasizes genetic origin and sequence similarities (both nucleotide and amino acid) rather than protein product size. For this purpose, the gene for the 43-kD protein from the heart (Beyer et al., 1987) has been designated as the c~ GJ gene, because the sequence for this protein is significantly different from those for the 32and 26-kD proteins from the liver (Paul, 1986; Kumar and Gilula, 1986; Nicholson and Zhang, 1988). The sequences for the other two proteins are quite similar, and therefore have been designated as members of the fl class of GJ genes: the 32-kD gene is referred to as the fl~ gene and the 26-kD gene is referred to as the ~2 gene (see Table I). In this study we have examined the expression of three different GJ transcripts (c~, Bj,/32) and their related protein products in various tissues during different stages of pregnancy to determine if there is a modulation of GJ expression in response to hormonal stimuli during pregnancy. Materials and Methods I. Animals and Tissue Collection Timed pregnant Wistar rats (220-250 g body weight), with a gestational period of 22 d, were obtained from Simonscn Laboratories (Gilroy, CA). Day 0 of pregnancy was defined by the presence of a uterine plug. The aniTable I. Gap Junction Genes and Related Products Deduced Reported protein mRNA protein product sizes Gene size product size (Cx) Reference*