Objective
Endogenous uterine agonists can activate numerous signaling pathways to effect increased force. Our objective was to assess expression of key constituents of these pathways, in alliance with contractile function, through late gestation and during term and preterm labor.
Study Design
Using myography, we measured the response to 3 agonists compared with depolarization alone (K + , 124 mEq/L) and calculated agonist/depolarization ratio. We measured gene expression using quantitative reverse transcription–polymerase chain reaction.
Results
Contractile responsiveness to depolarization alone, oxytocin, or endothelin-1 increased during pregnancy compared with nonpregnant animals. The agonist/depolarization ratio did not change during uterine activation or parturition. Inhibition of rhoA-associated kinase decreased responses to oxytocin in all tissues, but significantly more during uterine activation. Expression of rhoA and rhoA-associated kinase was increased significantly in active labor at term or preterm.
Conclusion
The rhoA/rhoA-associated kinase pathway is a key regulator of uterine activation during labor and may be a useful target for the prevention of spontaneous preterm birth.
More than 70% of newborn infant death and long-term developmental disabilities are associated with preterm birth. These adverse birth outcomes result in tremendous fiscal, personal, and societal costs. Our lack of understanding of the regulation of the timing of parturition underlies the failure to diminish the incidence of preterm birth over the past several decades.
Endogenous uterine agonists stimulate G-protein coupled receptors to activate 2 major signaling pathways in uterine myocytes ( Figure 1 ), as reviewed in Aguilar and Mitchell. First, the phospholipase C-mediated pathway leads to depolarization and subsequent Ca 2+ -dependent activation of myosin light chain kinase (MLCK). This results in phosphorylation of myosin regulatory light chain (RLC phosphorylated to pRLC), which triggers the contraction. The second pathway involves the activation of the monomeric G-protein, rhoA, with subsequent activation of its major effector, rhoA-associated kinase (ROK). A major effect of ROK activation is the inhibition of myosin RLC phosphatase (MLCP), which potentiates the pro-contractile effects of phosphorylated RLC. Additionally, the catalytic subunit of MLCP can be inhibited directly by small molecules such as the 17-kilodalton protein kinase C–potentiated inhibitor of protein phosphatase 1c (CPI-17) and phosphatase holoenzyme inhibitor (PHI) Both of these inhibitors can be activated by protein kinase C (PKC) or ROK.
Before parturition, the myometrium undergoes a process termed uterine activation , which prepares it for labor. This process remains poorly understood, despite the likelihood that many, if not most, cases of preterm birth result from a variety of pathologic events that cause early implementation of the process of uterine activation. The pathways noted in Figure 1 have been studied in a variety of tissues and species. It has been well-established that the activation of MLCK is dependent on an increase in cytosolic Ca 2+ , whereas the regulation of MLCP by ROK is essentially independent of Ca 2+ . Physiologic uterine agonists activate both the Ca 2+ -dependent and the Ca 2+ -independent pathways. In contrast, depolarization of the myocyte membrane predominantly activates only the Ca 2+ -dependent pathway. Thus, the difference between the 2 is largely dependent on the rhoA/ROK pathway. Comparison of the 2 types of stimulation provides information about each of these 2 distinct pathways. The objectives of the current studies were to obtain a systematic and integrated portrayal of these events in a single species with the use of a combination of in vivo, ex vivo, and in vitro methods to assess functional and biochemical endpoints. We have chosen the rat model because we have demonstrated that it shares with humans common molecular mechanisms that are involved in uterine contractile quiescence and activation. We have examined the effects in uterine tissues of the endogenous agonists oxytocin, endothelin-1, and prostaglandin F 2α (PGF 2α ) by measuring force generation of uterine strips in a muscle bath apparatus and gene expression of several key components of these contractile pathways in uterine tissues.
Materials and Methods
Animals
The University of Alberta Health Sciences Animal Policy and Welfare Committee approved all animal protocols, and the experiments were conducted in accordance with the Guidelines and Policies of the Canadian Council on Animal Care. Time-mated animals were obtained from the University of Alberta Health Sciences Laboratory Animal Services. Normal parturition occurs on the afternoon of gestation day 21 or on day 22. Animals were killed with a a lethal dose of isofluorane (Bimeda-MTC, Cambridge, Ontario, Canada). Study groups included 6 nonpregnant animals, 25 animals in mid-late pregnancy (gestation day 16-20), and 19 animals at gestation day 21 or 22. Of the latter group, there were no signs of labor (NIL) in 12 animals, and 7 animals were in active labor (LAB, defined as after delivery of the first pup).
As a model of preterm birth, an additional group of animals (n = 21) received a single subcutaneous injection of the progesterone receptor antagonist RU486 (2.5 mg dissolved in corn oil on the morning of gestation day 15; controls (n = 6) received a corresponding amount of corn oil) on the morning of gestation day 15. These animals were killed for tissue collection after 24 hours on gestation day 16. We have previously demonstrated that this treatment results in preterm birth in an average of 27 hours after injection. In all RU486-treated animals, there was evidence of LAB at the time of delivery (vaginal bleeding or delivery of a pup), whereas placebo-treated animals showed no such signs. The uterus was removed immediately after death. After removal of the pups, placentas, and fetal membranes, 1 horn of the uterus was frozen and stored at –80°C, and the other horn was used to obtain full-thickness strips of uterine wall for the myographic studies.
Quantitative reverse transcription-polymerase chain reaction (PCR)
These procedures were performed by standard procedures in our laboratory as described. Briefly, RNA was extracted with the GenElute Mammalian Total RNA Miniprep kit (Sigma-Aldrich, St. Louis, MO). Reverse transcription was performed with the Taqman Reverse Transcription Reagents kit (Applied Biosystems, Foster City, CA). The resultant complementary DNA (cDNA) was stored at –20°C until PCR.
Real-time PCR was performed in triplicate with the SYBR Green PCR Core Reagents Kit (Applied Biosystems) in an iCycler apparatus (Bio-Rad Laboratories, Ltd, Montreal, Quebec, Canada). The forward and reverse primer sequences, gene accession numbers, and amplicon locations within the gene are presented in the Table . Blank controls consisted of reactions in which sterile water was substituted for the cDNA template. The resulting SYBR-green fluorescence was measured at the end of each PCR cycle and plotted against the cycle number to determine the threshold cycle. After each PCR reaction, the purity of the amplified cDNA was confirmed by the presence of a single peak in the melt curve for all samples. For quantification, we used the “Δ-Δ” method to determine in relative expression between groups.
| Variable | Accession no. | Forward primer | Reverse primer | Amplicon (base pair) |
|---|---|---|---|---|
| CPI-17 | AF352572 | CGAGTCACCGTCAAATACGACC | GACCTCATCTGGCATGTCTGC | 120 |
| Cyclophilin | NM017101 | CACCGTGTTCTTCGACATCAC | CCAGTGCTCAGAGCTCGAAAG | 114 |
| Extracellular signal-regulated kinase–1 | AF155236 | AGGAGCCATTCACCTTTGAC | GGGTGTCTGTTCTTGTTAGGG | 128 |
| Extracellular signal-regulated kinase–2 | M64300 | TAAATTGGTCAGGACAAGGGCTCA | CCTCGGAACGGCTCAAAGGA | 132 |
| Integrin-linked kinase | NM133409 | CGCACTCAATAGCCGTAG | CTGAGCGTCTGTTTGTATCT | 164 |
| Phosphatase holoenzyme inhibitor | NM172045 | CTGTTCCCTCGCCCTCA | ATACAGCCTTTGGAGC | 134 |
| RhoA | AY026068 | TGCCCATCATCCTAGTTG | CTCCGTCTTTGGTCTTTGCT | 171 |
| Rho-associated kinase–1 | NM031098 | CGGCTAACTGACAAGCATCA | TCAGCTTCTCTCGAGCTTC | 104 |
| Rho-associated kinase–2 | U38481 | AGATCAGTGCAGCGGCTATT | TCACTACCACGCTTGACAGG | 130 |
| Rnd-1 | BC089994 | GAGCAGGGCTGTGCGATA | GCTTCGGACAGGGCTCTT | 159 |
| Rnd-2 | BC089058 | ACATGCGGACTGACCTGG | TGGAAGACGTCCCTGACA | 160 |
| Rnd-3 | NM001007641 | AAATACCAGGCAGACTCC | TGACGCTGTTCTCCGACT | 110 |
Uterine myography
The muscle bath tension experiments were performed as described in published methods. Briefly, the uterine horn was incised longitudinally, and full-thickness strips approximately 8 mm in length and 3 mm in width were excised, avoiding implantation sites. A resting tension of 1 g was used in all experiments. To confirm the physiologic viability that included sensitivity and responsiveness for each uterine strip, a cumulative concentration-response curve was established for oxytocin (0.1-320 nmol/L added at 3-minute intervals). The concentration at which half-maximal tension occurred was calculated as a measure of sensitivity for each of the contractile stimulants with Prism software (version 4.0; GraphPad Software, San Diego, CA). Only those strips that exhibited high correlation (r 2 > 0.80) for the curve fit were used for the data that are reported.
To compare the maximal contractile activity that is produced by the contractile agonists with results from high K + -induced depolarization, we measured the responses to oxytocin (320 nmol/L), endothelin-1 (1 μmol/L) and PGF 2α (10 μmol/L) compared with a high K + solution (124 mEq/L). To evaluate the responses to the 2 types of stimuli, we determined the agonist/depolarization response ratio (ADR). The contractile agonists were chosen on the basis of our previous studies. The integrated responses over 5 minutes after the addition of the stimulant ( Figure 2 , A) were compared after the subtraction of the baseline tension. Because these high concentrations of agonists elicit tonic responses, we chose the area under the curve as the appropriate parameter to assess responses. For the tissues that were obtained from the rats in the LAB group, only oxytocin was studied as a contractile agonist. In instances in which >1 strip from 1 animal was used, the data were averaged to provide a single estimate per animal before statistical evaluation.
To assess the effects of ROK or PKC inhibition, muscle strips were treated with oxytocin (5 nmol/L) with or without a 15-minute preincubation with H-1152 (10 –7 to 10 –5 mol/L) or bisindolylmaleimide (10 -9 – 10 -6 mol/L). For each dose of drug (or control), there was a 20-minute protocol during which the drug was incubated with the strip for 15 minutes then treated with oxytocin (5 nmol/L); the response was measured in the 5-minute interval immediately after the oxytocin. The stability of contractile pattern with 5 nmol/L oxytocin over this time period in the untreated control strips is shown in Figure 2 , B. The strips were washed thoroughly, and the protocol was repeated for the next dose of drug (or control). For analyses, the net integrated response to oxytocin (over 5 minutes, baseline activity subtracted) in the presence of the drug was expressed as a percent of the corresponding oxytocin-only control response that was determined immediately before the addition of the first concentration of drug.
Statistics
Statistical analyses were performed with InStat software (version 3.0; GraphPad Software) and Prism software. Statistical significance was considered at a probability value of ≤ .05. Comparisons of the responses to K + and the uterine contractile agonists for the nonpregnant and pregnant groups at different gestational ages were examined with 2-factor analyses of variance (ANOVA). Comparisons for each agonist in each separate group were then assessed with the use of a 2-tailed paired t test. The RU486-treated animals were compared with controls with the use of a 2-tailed unpaired t test. Comparisons among the 4 gestational age groups were assessed with the use of 1-factor ANOVA with post hoc Tukey-Kramer test when the probability value was < .05. Assessment of the pharmacologic agents was performed with the use of a 2-factor ANOVA with Bonferroni’s post-hoc test, if applicable.
Results
Changes in myographic assessment
The myographic data that compared the tensions generated by uterine agonists with those by K + are presented in Figure 3 . These agonists were selected because of their known stimulatory effects on the myometrium and because each is synthesized within the rat uterus, which suggests that they might have physiologic influence on myometrial contractility. The nonpregnant animals were used as a reference point. The data from individual animals are presented in Figure 3 , A; the means and standard errors for the grouped data are shown in Figure 3 , B. The data for each agonist in Figure 3 , B, were analyzed with the use of 2-factor ANOVA (variables were gestational age and treatment). These data revealed a statistically significant increase with oxytocin and endothelin-1 ( P < .0001), but a significant decrease ( P < .03) with PGF 2α . The effect of agonist treatment compared with K + was significant only for oxytocin ( P = .016); this was apparent at each gestational group ( Figure 3 , B). Only treatment with PGF 2α exhibited a significant interaction between the agonist and gestational grouping ( P = .027) where the response to PGF 2α was significantly lower than to K + in the tissues from rats at term.
When the data in Figure 3 , B, were analyzed separately for K + and for each of the agonists with 1-factor ANOVA, there was a significant change in the contractile responses of the strips to K + ( P = .019), oxytocin ( P = .003), and endothelin-1 ( P = .002), but not to PGF 2α . The peak responses in contractility occurred in the group at gestation day 21/22 NIL. For the tissues that were obtained during LAB, we tested only the response to oxytocin, which was significantly less in comparison with the NIL group.
Because we were interested in comparing the responses of a physiologic agonist to depolarization alone (with K + ), we calculated the agonist to depolarization ratio (ADR) for each group ( Figure 3 , C). The ADR was reduced in all groups during pregnancy compared with the nonpregnant controls. The ADR for oxytocin was 2.2 ± 0.3 in strips from nonpregnant rats and was diminished during gestation (to 1.5 ± 0.5, 1.4 ± 0.4, and 1.3 ± 0.1 in the gestation day 18/19 and 20/21 NIL and the LAB group, respectively). Thus, despite the increased responses in contractility with either K + or endogenous agonist as demonstrated in Figure 3 , B, the proportion because of agonist-dependent mechanisms is not greater than K + .
We further explored potential mechanisms that regulate the responses to oxytocin using pharmacologic inhibitors that target ROK and PKC ( Figure 4 ). The responses of the strips to repeated challenges with 5 nmol/L oxytocin remained remarkably stable through the experimental protocol. The highly specific ROK inhibitor H-1152 caused a significant ( P < .001, 2-factor ANOVA) concentration-dependent reduction in the response to oxytocin in strips from both nonpregnant ( Figure 4 , A) and pregnant animals at gestation day 18/19 or gestation day 21/22 NIL ( Figure 4 , B and C, respectively). The effect of H-1152 was significantly affected by both pregnancy and gestational age ( P = .016). The inhibition of oxytocin-stimulated activity was highest in the nonpregnant group; at the 2 highest doses, this was significantly greater inhibition than was seen in the tissues from the gestation day 18/19 animals. The tissues from the gestation day 21/22 animals were not significantly different from either of the other 2 groups. The PKC antagonist bisindolylmaleimide did not affect oxytocin-induced tension development in any of the groups.
