Functional Coupling of Rat Myometrial α1-Adrenergic Receptors to Ghα/Tissue Transglutaminase 2 during Pregnancy

Ghα protein, which exhibits both transglutaminase and GTPase activities, represents a new class of GTP-binding proteins. In the present study, we characterized Ghα in rat uterine smooth muscle (myometrium) and followed its expression during pregnancy by reverse transcription-PCR and Western blot. We also measured transglutaminase and GTP binding functions and used a smooth muscle cell line to evaluate the role of Ghα in cell proliferation. The results show that pregnancy is associated with an up-regulation of Ghα expression at both the mRNA and protein level. Ghα induced during pregnancy is preferentially localized to the plasma membrane. This was found associated with an increased ability of plasma membrane preparations to catalyze Ca2+-dependent incorporation of [3H]putrescine into casein in vitro. In the cytosol, significant changes in the level of immunodetected Ghα and transglutaminase activity were seen only at term. Activation of α1-adrenergic receptors (α1-AR) enhanced photoaffinity labeling of plasma membrane Ghα. Moreover, the level of α1-AR-coupled Ghα increased progressively with pregnancy, which parallels the active period of myometrial cell proliferation. Overexpression of wild type Ghα in smooth muscle cell line DDT1-MF2 increased α1-AR-induced [3H]thymidine incorporation. A similar response was obtained in cells expressing the transglutaminase inactive mutant (C277S) of Ghα. Together, these findings underscore the role of Ghα as signal transducer of α1-AR-induced smooth muscle cell proliferation. In this context, pregnant rat myometrium provides an interesting physiological model to study the mechanisms underlying the regulation of the GTPase function of Ghα

Different from the traditional heterotrimeric and monomeric G proteins, Gh␣ protein is a bifunctional enzyme with both transglutaminase and GTPase activities (1). Transglutamination requires Ca 2ϩ and is inhibited by GTP, whereas GTPase activity is inhibited by Ca 2ϩ . Initially identified as tissue transglutaminase 2 (TG2), 1 Gh␣ is a member of a large family of transglutaminases that includes plasma Factor XIIIa, epidermal TG1 and TG3, or prostate TG4 (2). These enzymes catalyze the formation of covalent ␥-glutamyl-⑀-lysine bonds between proteins or polyamines and protein substrates. Such modifications play a role in different biological processes such as blood coagulation, epidermal differentiation, formation of copulatory plug in rodents, or extracellular matrix organization (2). Like other transglutaminases, Gh␣ consists of four domains: an amino-terminal ␤-sandwich and a catalytic core with the active site cysteine (Cys-277) for transglutamination, followed by two carboxyl-terminal ␤-barrels. The three-dimensional structure of Gh␣ complexed with GDP revealed a unique guanine nucleotide binding pocket located between the catalytic core and the first ␤-barrel and formed by residues coming from both domains (3). By virtue of its GTP binding/GTPase activity, Gh␣ acts as a signaling molecule for ␣1-AR (4,5), oxytocin receptors (6,7), and thromboxane receptors (8). ␣1-AR were, however, the most studied Gh␣ protein-coupled receptors. Activation by ␣1-AR induces exchange of GDP to GTP and dissociation of GTP-Gh␣ from the Gh␤ subunit, which was recently identified as calreticulin (9). GTP-bound Gh␣ interacts with downstream effector PLC␦1, thereby resulting in phosphoinositide hydrolysis and Ca 2ϩ increase (10 -13). Deactivation of this signaling pathway is triggered by GTP hydrolysis and reassociation of Gh␣ with free Gh␤ subunit. Although ␣1-AR/Gh␣ coupling was well established in many cell types, the cellular responses triggered by this signaling pathway still remain unclear. Indeed, only a few studies have addressed the role of Gh␣ as a signal transducer of G protein-coupled receptors.
The present study was undertaken to investigate the potential involvement of Gh␣ in myometrial ␣1-adrenergic-induced responses. It is well known that norepinephrine modulates myometrial contractility during pregnancy. Adrenergic signaling pathways are under the control of progesterone and estradiol, which regulate the expression of receptors (␤-AR), heterotrimeric G proteins (Gs␣, Gi␣, and Gq␣), and PLC␤ enzymes (14 -17). Hence, during pregnancy under progesterone dominance, norepinephrine stimulates ␤-AR and ␣2-AR, which both activate adenylyl cyclase and increase intracellular concentrations of cAMP (18,19). The latter second messenger induces myometrial relaxation by inhibiting the pathways leading to Ca 2ϩ increase (20). At term, when concentrations of estradiol rise, the ␤-adrenergic pathway becomes desensitized (21), and norepinephrine acts on ␣2-AR (19,22) and ␣1-AR (23). At this time, ␣2-AR shift to inhibitors of adenylyl cyclase, whereas ␣1-AR activate the Gq/PLC␤ system and participate in the Ca 2ϩ increase and uterine contraction. An intriguing observation is that ␣1-AR are expressed throughout pregnancy, although they are not efficiently coupled to phosphoinositide hydrolysis due to down-regulation of Gq␣ (16) and PLC␤ isoforms at this period (17). We thus questioned whether these receptors could signal through Gh␣ to regulate response(s) other than contraction. Indeed, myometrial cells undergo proliferation during pregnancy, which is critical for uterine adap-tation to fetoplacental growth (24,25). Although many studies have addressed the regulation of such a process in physiopathological situations like uterine leiomyoma, very little is known about the physiological processes involved during pregnancy.
For this purpose, we first characterized myometrial Gh␣ and followed its expression, at the mRNA and protein level, during pregnancy and at term. We also determined myometrial transglutaminase activity and photoaffinity labeling of Gh␣ in the absence or presence of GTP or ␣1-adrenergic agonist. Finally, we used a smooth muscle cell line to analyze the role of Gh␣ and its transglutaminase-inactive (C277S) mutant in cell proliferation induced by ␣1-AR. Animals and Tissues-Sprague-Dawley rats were obtained from Janvier (Le Genest, France). The females were caged with males overnight, and successful mating was determined by the presence of spermatozoa in the vaginal smear (day 1 of pregnancy). Animals were sacrificed by cervical dislocation at days 5, 12, 15, and 20 of pregnancy or at term during the expulsion of fetoplacental units, following the guidelines laid down by the National Institutes of Health guide. The uterine horns were quickly isolated and cut open lengthwise, and the fetoplacental units were removed. The myometrium was then freed of adherent endometrium.

Materials-[
Cell Culture and [ 3 H]Thymidine Incorporation-DDT1-MF2 cells were grown in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum and 25 units/ml penicillin and streptomycin. For [ 3 H]thymidine uptake, cells were seeded in 12-well plates and transfected (0.6 g of DNA/0.12 ϫ 10 6 cells/well) with expressing vector for wild type or mutant (C277S) Gh␣ constructs using LipofectAMINE Reagent Plus. The cells transfected with the empty pcDNA vector were used as control. DDT1-MF2 cells were then given fresh serum-free medium and allowed to achieve quiescence for 24 h prior to subsequent application of 10 M phenylephrine for an additional 24 h. DNA synthesis was assessed following the addition of [ 3 H]thymidine (1 Ci/ml) for a period of 6 h before the end of the treatment protocol. Cells were then washed with cold phosphate-buffered saline, and cold 5% trichloroacetic acid was added to precipitate DNA as described previously (26). The precipitates were resuspended in 0.5 N NaOH, and aliquots were counted by scintillation counting. Experiments were done in triplicate. For Western blot analysis, DDT1-MF2 cells were plated in 6-well plates (0.3 ϫ 10 6 cell/well), and the transfected DNA of the empty vector, wild type, or mutant Gh␣ was 1.2 g of DNA/well. RNA Preparation and Reverse Transcription (RT)-PCR-Total RNA was extracted from rat myometrium and liver as described previously (27). The RT-PCR reactions were done with the kit from Invitrogen. Briefly, 3 g of total RNA were reverse transcribed according to the instructions of the manufacturer, and the resulting cDNAs were stocked at Ϫ80°C. A 0.05 volume of each RT reaction was amplified using specific upstream and downstream primers for Gh␣ and the internal control glyceraldehyde-3-phosphate dehydrogenase (GAPDH). Primers sequences for rat Gh␣ (8) and rat GAPDH (28) were 5Ј-ATC-TACCAGGGCTCTGTCAA-3Ј and 5Ј-ACTCCACCCAGCAGTGGAAA-3Ј, 5Ј-CCATGGAGAAGGCTGGGG-3Ј and 5Ј-CAAAGTTGTCATGGAT-GACC-3Ј, respectively. PCR reactions (40 cycles) were 94°C for 2 min, 56°C for 1 min, and 72°C for 2 min, with initial activation of the enzyme at 94°C for 5 min. For semiquantitative analysis, PCR cycle profiles were conducted, and the cycle number (25 cycles) was chosen from the linear portion of the curve. The resulting products were run on 2% agarose gel and quantitated by ethidium bromide fluorescence.
Preparation of Myometrial Fractions-Rat myometrium and liver were homogenized in buffer A for Western blot analysis, in buffer B for transglutaminase assay, and in buffer C for photoaffinity labeling experiments. After 10-min centrifugation at 4°C, supernatants were collected and submitted to 100,000 ϫ g centrifugation at 4°C for 1 h to separate plasma membranes from cytosol. Pellet containing plasma membranes was resuspended in homogenization buffer, and protein concentration of plasma membrane and cytosolic fractions was determined according to Bradford (29) with bovine serum albumin as standard. Samples were stored at Ϫ80°C until use.
Western Blot Analysis-Proteins from rat myometrium and liver (20 g of protein) or DDT1-MF2 cells (5 g of protein) were separated by 7.5% SDS-polyacrylamide gels and transferred to PVDF membranes. The blots were blocked overnight at 4°C in Tris-buffered saline containing 5% nonfat dried milk and were incubated for 1 h at room temperature with monoclonal anti-Gh␣ or anti-PLC␦1 diluted 1:500. Incubation with secondary antibody (1:10,000) was carried out for 45 min at room temperature. Immunoreactive bands were visualized by the chemiluminescence detection system and quantified. The quantification of Gh␣ and PLC␦1 expression was determined by densitometric scanning followed by computer analysis using the NIH Image 1.62 program.
Transglutaminase Assay-The ability of plasma membrane and cytosolic fractions to catalyze the incorporation of [ 3 H]putrescine into dimethylcasein was determined as described previously (30). Briefly, 15 g of proteins were incubated with 0.8% (w/v) dimethylcasein and 10 M [ 3 H]putrescine in a 50-l total volume of buffer containing 40 mM Tris, pH 7.4, 10 mM MgCl 2 , 30 mM dithiothreitol, 0.4 mM EDTA, 0.2 mM EGTA, 20% glycerol. Reactions were initiated at 37°C in the absence or presence of 2 mM CaCl 2 or 0.5 mM GTP. After 40-min incubation, reactions were stopped by addition of 50% cold trichloroacetic acid. The pellets were washed three times with cold 10% trichloroacetic acid and solubilized with 0.1 M NaOH. Radiolabeling was measured by scintillation counting.
Photoaffinity Labeling-Photoaffinity labeling of Gh␣ with [␣-32 P]GTP was carried out as described previously (4). Plasma membrane fractions (200 g of protein) were incubated with 10 Ci of [␣-32 P]GTP for 10 min at 30°C in buffer containing 20 mM Hepes, pH 7.5, 1 mM EGTA, 0.5 mM dithiothreitol, 10% glycerol, 100 mM NaCl, 2 mM MgCl 2 , 0.5 mM AppNHp in the absence or presence of 10 M phenylephrine, 0.5 mM unlabeled GTP, or 100 M phentolamine. Reactions were stopped in ice; the samples were irradiated with UV light (254 nm) for 20 min and solubilized with Laemmli solution for 1 h at room temperature. The samples were then subjected to SDS-PAGE in 7.5% gels and transferred to PVDF membranes. The membranes were autoradiographed and then immunodetected for Gh␣. The levels of labeled and immunodetected Gh␣ were determined by densitometric scanning followed by computer analysis using NIH Image 1.62 program.
Statistical Analysis-Results were expressed as means Ϯ S.E. Statistical significance was assessed by Student's t test for unpaired data. A probability level less than 0.05 was considered statistically significant.

Characterization of Rat Myometrial
Gh␣-To assess the expression of Gh␣ in pregnant rat myometrium, we used RT-PCR technique and Western blot analysis. Rat liver where the expression of Gh␣ has been reported already was used as control. Using specific primers, we amplified a 520-bp fragment of Gh␣ from pregnant rat myometrium and rat liver (Fig. 1A, lanes 2  and 4). Control PCR reactions performed on nontranscribed RNAs indicated no contamination of the RNA preparations with genomic DNA (Fig. 1A, lanes 1 and 3). Monoclonal antibodies raised against Gh␣ stained a single band of the appropriate molecular mass (74 kDa) in plasma membranes and cytosolic fractions of pregnant rat myometrium and rat liver (Fig. 1B). Our results also show that the subcellular distribution of Gh␣ is tissue-specific (Fig. 1B). Indeed, Gh␣ is predominantly cytosolic in liver, while a high amount of this protein is associated with the myometrial plasma membranes.
Rat Myometrial Gh␣ during Pregnancy-We examined the expression level of myometrial Gh␣ at different stages of pregnancy. Semiquantitative analysis showed that the level of Gh␣ transcripts, but not of the internal control GAPDH, is altered during pregnancy ( Fig. 2A). Indeed, mRNA level of Gh␣ increased progressively during pregnancy, with pronounced changes occurring at day 12 of pregnancy (2-fold of day 5) and at term (1.5-fold of day 20) (Fig. 2B). Quantification of the immunodetected Gh␣ also indicated an up-regulation of this protein. Indeed, the amount of membrane-associated Gh␣ increased progressively during pregnancy (Fig. 3A). In contrast, the amount of cytosolic Gh␣ remained stable until day 20 of pregnancy ( Fig. 3B), and a significant elevation (ϩ35%) was seen only at term. Comparison of the subcellular distribution of Gh␣ during pregnancy was expressed as the membrane/cytoplasmic ratio. The obtained data illustrated in Fig. 3C revealed a progressive enrichment of plasma membrane compartment in Gh␣ during the course of pregnancy. Indeed, whereas the cytosolic localization of Gh␣ prevailed over membrane-associated Gh␣ at day 5 of pregnancy, this situation was completely reversed at the end of pregnancy (ratio of 0.56 at day 5 versus 1.5-1.8 between day 20 and term). Altogether, these results indicate an up-regulation of rat myometrial Gh␣ during pregnancy. Moreover, this period is associated with a preferential increase of membrane-bound Gh␣.
Myomerial Transglutaminase Activity during Pregnancy-Gh␣ is a bifunctional enzyme with both transglutaminase and GTPase activities. To characterize myometrial transglutaminase activity, we tested the ability of plasma membrane and cytosolic fractions to catalyze, in vitro, the incorporation of [ 3 H]putrescine into dimethylcasein. Addition of 2 mM CaCl 2 increased transglutaminase activity in both plasma membrane and cytosolic fractions of pregnant rat myometrium (6.4-fold and 4.8-fold above basal in the plasma membrane and the cytosol, respectively) (Fig. 4A). To evaluate the degree of participation of Gh␣ in this myometrial activity, we tested the inhibitory effect of GTP. As shown in Fig. 4A, 0.5 mM GTP blocked ϳ90% of Ca 2ϩ -stimulated transglutaminase activity in plasma membrane and cytosolic compartments.
We also examined myometrial transglutaminase activity at different stages of pregnancy. In the absence of Ca 2ϩ , plasma membrane and cytosolic transglutaminase activity did not change throughout pregnancy (a mean of 121 and 66 fmol/mg of protein/min in plasma membrane and cytosol, respectively). Addition of Ca 2ϩ significantly increased transglutaminase activity, and GTP counteracted this elevation in both fractions at all stages of pregnancy studied (data not shown). As illustrated in Fig. 4, B and C, Ca 2ϩ -stimulated activity of the plasma membrane fraction increased progressively during pregnancy, whereas the cytosolic transglutaminase activity was stable un-til day 20 of pregnancy. In correlation with Western blot analysis, a significant increase (ϩ30%) of the cytosolic transglutaminase activity was observed only at term (Fig. 4C).
Functional Coupling of Myometrial ␣1-AR to Gh␣-We next analyzed whether myometrial Gh␣, by virtue of its GTP binding activity, interacts with ␣1-AR. For this purpose, plasma membrane preparations were incubated with [␣-32 P]GTP in the absence or presence of 10 M phenylephrine (␣1-adrenergic agonist). Labeled G proteins were then separated by electrophoresis and transferred to PVDF membranes prior to autoradiography and immunodetection. The molecular mass of the 74-kDa radiolabeled protein coincided with that of the immunoreactive Gh␣ detected by Western blot analysis (Fig. 5A, top  and bottom). Addition of unlabeled GTP completely blocked labeling of Gh␣ (Fig. 5A, top). Moreover, application of phenylephrine significantly augmented photoaffinity labeling of Gh␣ (140 -160% of basal) at all stages of pregnancy studied (Fig. 5,  A (top) and B). These responses were blocked by addition of phentolamine, an ␣1-adrenergic antagonist (data not shown). Interestingly, both basal and phenylephrine-stimulated [␣-32 P]GTP labeling of Gh␣ increased during pregnancy (Fig.  5B). The maximal amounts were observed between day 20 of pregnancy and term. In this comparative study, labeled Gh␣ was hardly detectable at day 5 of pregnancy, even in the presence of phenylephrine.
Gh␣ in DDT1-MF2 Smooth Muscle Cell Line-␣1-AR/Gh␣ coupling was shown to activate PLC␦1, thereby resulting in phosphoinositide hydrolysis and Ca 2ϩ increase (11,13). To investigate whether such activation occurs in pregnant rat myometrium, we first characterized the expression of PLC␦1. Results illustrated in Fig. 6A indicated the presence of this PLC enzyme in non-pregnant, pregnant, and term rat myometrium. However, the amount of PLC␦1 during pregnancy represents only 30% of that observed in term rat animals (Fig. 6B). Pregnancy-dependent down-regulation of PLC␦1 is in correlation with the fact that all pathways leading to Ca 2ϩ increase and contraction are inhibited or reduced during pregnancy. We thus asked whether ␣1-AR/Gh␣ coupling could be involved in myometrial proliferation. To answer this question, we used the DDT1-MF2 cell line, which was described previously as a useful model to study the myometrial adrenergic signaling pathway regulations (31, 32). As myometrial cells, this smooth muscle cell line expresses endogenously ␣1-AR (33). Transfection of DDT1-MF2 cells with cDNAs encoding wild type or transglutaminase-inactive (C277S) Gh␣ successfully resulted in an increase of the amount of Gh␣ (Fig. 7A). Assessment of transglutaminase activity showed that addition of 2 mM CaCl 2 enhanced in vitro [ 3 H]putrescine incorporation into casein only in extracts derived from wild type-transfected cells, thus confirming the lack of this function in C277S-expressing cells (Fig.  7B). To determine whether ␣1-AR stimulate DNA synthesis through Gh␣, serum-starved transfected cells were incubated in the absence or presence of phenylephrine. Fig. 7C shows that activation of ␣1-AR resulted in an increase of [ 3 H]thymidine incorporation into control cells (ϩ50% above unstimulated cells). Overexpression of wild type Gh␣ further increased phenylephrine-stimulated DNA synthesis (a mean of 90% above stimulated control cells). Interestingly, a similar response was observed in cells expressing C277S-Gh␣ (Fig. 7C). Therefore, these results indicate that Gh␣, via its GTP binding/GTPase function, participates in smooth muscle proliferation induced by ␣1-AR.

DISCUSSION
Expression of Gh␣ in Rat Myometrium-By using RT-PCR and Western blot analysis, we showed the expression of Gh␣ in pregnant and term rat myometrium. Transglutaminase assay also indicated the presence of a Ca 2ϩ -and GTP-sensitive transglutaminase. In addition, GTP inhibited ϳ90% of myometrial transglutaminase activity. Among members of the transglutaminase family, TG3 is also sensitive to guanine nucleotides (34). However, this transglutaminase presents a restricted expression to epiderm (2). Based on these observations, we can then easily conclude that Gh␣ is predominantly expressed in pregnant and term rat myometrium. Another argument in favor of the predominant expression of Gh␣ in pregnant rat myometrium is the good correlation among the levels of immunodetected Gh␣ protein, myometrial transglutaminase activ- ity, and photoaffinity-labeled Gh␣ during pregnancy. These findings could explain the absence of additional bands in rat myometrium compared with mouse heart (35) by using the NeoMarkers monoclonal antibodies.
The presence of Gh␣ was also reported in non-pregnant human myometrium (6). However, its regulation has not been studied throughout pregnancy. The present study shows that the mRNA and protein levels of Gh␣ increase with pregnancy. This suggests that myometrial Gh␣ gene is regulated, at least in part, at the transcriptional level. Our preliminary experiments indicate that myometrial Gh␣ is not under the control of progesterone or estradiol (data not shown). A possible candidate for such a regulation could be retinoic acid. Indeed, this potent inducer of Gh␣ expression, as shown in many cell types, is locally synthesized and stored in pregnant rat uterus (36,37). In addition, rat myometrium expresses different retinoic acid receptors (RAR␣, RAR␤, and RXR␤) that could transduce retinoic acid-induced effects (37,38). Future studies will examine whether myometrial Gh␣ is regulated by retinoic acid during pregnancy. Preferential Association of Induced Gh␣ with Myometrial Plasma Membranes during Pregnancy-Gh␣ is known to reside predominantly in the cytosol. However, some studies localized a portion of this protein in the plasma membrane compartment (2). The present study reports a progressive enrichment of myometrial plasma membranes in Gh␣ during pregnancy. Indeed, Western blot analysis and transglutaminase assay as well as photoaffinity labeling study showed an increase of plasma membrane-associated Gh␣, whereas the amount of Gh␣ in the cytosol remained unchanged until day 20 of pregnancy. The molecular mechanisms underlying this differential localization of intracellular Gh␣ during pregnancy need to be clarified. It is possible that Gh␣ actively translocates from the cytosol to the plasma membrane during pregnancy because this protein was reported to moonlight between these compartments (39,40). Alternatively, retinoic acid was shown to increase the ability of Gh␣ to associate with the plasma membrane in HeLa cells (41).
Interaction between Myometrial ␣1-AR and Gh␣-By using photoaffinity labeling technique, we showed that myometrial Gh␣ functionally interacts with ␣1-AR in pregnant and term rats. The progressive enhancement in the level of ␣1-AR-coupled Gh␣ correlates well with the up-regulation of plasma membrane-associated Gh␣ during pregnancy. This strongly suggests that plasma membrane-associated Gh␣ is accessible to the interaction with G protein-coupled receptors in pregnant rat myometrium. Previous findings in human vascular smooth muscle cells reported that the particulate Gh␣ codistributes with stress fibers and may thus stabilize cytoskeletal structures through its cross-linking function (42). In the latter study, the particulate Gh␣ appeared inactive when assayed by in vitro putrescine/casein assay, maybe due to its tight binding to preferred local substrates (42). It is known that stimulation of transglutaminase activity into cells requires both high concentrations of Ca 2ϩ and a decrease of guanine nucleotide levels (43,44). In rat myometrium, pathways leading to Ca 2ϩ increase and uterine contraction are inhibited during pregnancy to allow development of fetoplacental units. In addition, in the presence of exogenous Ca 2ϩ , myometrial plasma membrane Gh␣ was able to catalyze the incorporation of putrescine into casein. These discrepancies between both models lead us to suggest that plasma membrane Gh␣ plays different roles in uterine and vascular smooth muscles.
The present work supports previous findings in that intracellular localization of Gh␣ dictates its functions. Indeed, Gh␣ purified from plasma membranes was shown to exhibit higher GTP binding activity than the cytosolic Gh␣ in mouse heart (45). In addition, translocation of Gh␣ from the plasma membrane to the cytosol is accompanied by the loss of GTP binding and the appearance of transglutaminase activity (39,40). In myometrium, the up-regulation of cytosolic Gh␣ at term coincides with the onset of labor. At this time, the very high elevation of intracellular concentrations of Ca 2ϩ could stimulate transglutaminase activity. Recent data from mice lacking TG2 showed that this function is important for the stabilization of apoptotic thymocytes before their clearance (46). Therefore, it is tempting to suggest that the cytosolic Gh␣ stabilizes dying myometrial cells during uterine involution that occurs after delivery.
Gh␣-induced Smooth Muscle Cell Proliferation-We have shown previously that myometrial ␣1-AR participate to the initiation of uterine contraction at term through activation of the Gq/PLC system (23,47). During pregnancy, Gq␣ and PLC␤ isoforms are down-regulated, thereby resulting in a weak phosphoinositide hydrolysis in response to the activation of ␣1-AR (16,17). The present study shows that ␣1-AR interact with Gh␣ in pregnant rat myometrium. Moreover, the level of ␣1-ARcoupled Gh␣ increased during pregnancy, with a peak reached at day 20. Interestingly, our recent findings reveal a similar pattern for the increase of the myometrial weight and DNA amount during pregnancy. 2 The correlation between ␣1-AR/ Gh␣ coupling and myometrial proliferation could be highly relevant to understanding the physiological significance of ␣1-AR expression during pregnancy, particularly given that ␣1-AR mediate proliferation of several smooth muscle cell types (48).
To address this question, we transfected DDT1-MF2 cells with wild type and transglutaminase-inactive Gh␣. Previous studies have shown that mutating the active site cysteine (Cys-277) impairs transglutaminase activity of Gh␣ without affecting its GTP binding/GTPase function and interaction with ␣1-AR (49). The results indicated that ␣1-AR operate through Gh␣ to stimulate smooth muscle cell proliferation. Firstly, overexpression of Gh␣ enhanced ␣1-adrenergic-induced DNA synthesis. Secondly, the use of C277S-Gh␣ showed that this response involves the GTP binding/GTPase function of Gh␣. Some studies in vascular smooth muscle cells reported the involvement of heterotrimeric G proteins in ␣1-AR-induced proliferation (48). In the present work, we describe an additional mechanism by which ␣1-AR could regulate smooth muscle cell proliferation. Indeed, our data are the first demonstration of the involvement of Gh␣ in such a cellular response. The molecular mechanisms underlying smooth muscle cell proliferation remain to be clarified. In rat hepatocytes, it was suggested that Gh␣ could act on PLC␦1 to induce cell proliferation (50). In non-pregnant human myometrium, such interaction between Gh␣ and PLC␦1 was described (7). Nevertheless, our results indicate that if PLC␦1 could be a downstream effector for ␣1-AR/Gh␣ coupling and participate in the Ca 2ϩ increase and uterine contractions at term, this seems unlikely during pregnancy. In fact, at this period, PLC␦1 is down-regulated, as are the majority of the contraction-associated proteins. Gh␣ was also shown to participate in the activation of extracellular signal-regulated kinases by ␣1-AR in neonatal rat cardiomyocytes (51). Further studies will define whether Gh␣ activates the mitogen-activated protein kinase pathway to induce myometrial cell proliferation during pregnancy.
In summary, our results reveal, for the first time, that the expression of Gh␣ is induced in pregnant rat myometrium. Moreover, the induced protein preferentially associates with the plasma membrane where it interacts with ␣1-AR. During pregnancy, Gh␣ may play an important role in the transduction of myometrial ␣1-adrenergic signaling. Firstly, we have shown previously that Gq␣ is down-regulated at this period. Secondly, Gh␣ enhanced ␣1-adrenergic-induced proliferation of DDT1-MF2 smooth muscle cell line.