Monitoring uterine contractility in mice using a transcervical intrauterine pressure catheter.

Proper uterine contractility is not only required for labor progression and the control of postpartum hemorrhage, but also for the transport of semen and gametes, as well as embryo implantation. Inadequate uterine contractility contributes to retrograde menstruation, endometriosis, infertility, ectopic pregnancies and miscarriages ( Lesny et al. 1999 , Bulletti et al. 2002 , Kissler et al. 2004 , Leyendecker et al. 2004 , Bulletti & de Ziegler 2005 , Aguilar & Mitchell 2010 ). Conversely, uterine hyper-contractility is a main cause for primary dysmenorrhea ( Aguilar & Mitchell 2010 ). In women, uterine activity (frequency, strength, and duration of contractions) during labor is monitored clinically by either: external palpation, electrohysterography (EHG; also known as uterine electromyography), magnetomyography ( Eswaran et al. 2002 ), external tocodynamometry, or internal tocodynamometry utilizing a transcervical intrauterine pressure catheter (IUPC) ( Pauli & Repke 2016 ). External tocodynamometry is useful for long-term, non-invasive monitoring of contraction frequency and duration; however, IUPC accurately monitors these contractile parameters as well as amplitude with minimal artifact. While IUPC is the gold standard for quantitatively measuring intrauterine pressure (IUP) and activity ( Hadar et al. 2015 ) in the United States, its use is typically restricted to high risk pregnancies to monitor labor progress ( Pauli & Repke 2016 ). The rationale behind the limited use of IUPC is risk of infection, due to its insertion into the amniotic space following rupture of the fetal membranes. It is estimated that approximately 20% of the deliveries in the United States are monitored using an IUPC ( Choby 2011 ). Rodent models are instrumental research tools to study pregnancy, preterm labor (PTL) and term labor, including, but not limited to: 1) transgenic and knockout mice used to decipher the role of a specific gene in mediating the timing of labor or mechanisms underlying uterine contractility, 2) determining the existence and location of a uterine pacemaker ( Fuchs & Poblete 1970 , Miller et al. 1989 , Lammers et al. 1994 , Lammers et al. 2015 ) and 3) pre-clinical testing of novel therapeutics to regulate uterine contractility for the prevention of PTL, labor induction, or control of postpartum hemorrhage. In rodent models utilized for pregnancy-related studies, uterine contractions or pressure have been measured by either ex vivo isometric tension, abdominal and uterine surface electromyography ( Buhimschi & Garfield 1996 , Buhimschi et al. 1998 ), or in vivo pressure recordings using either surgical placement of an IUPC ( Vickery 1979 , Buhimschi & Garfield 1996 , Buhimschi et al. 1998 , Mahendroo et al. 1999 , Word et al. 2005 ) or IUPC-radiotelemeter directly into the uterus ( Pierce et al. 2010 , Rada et al. 2015 ). Isometric tension recording is an excellent ex vivo functional assay to measure uterine myometrial contractility ( Pak et al. 1994 , Norman et al. 1997 , Baumbach et al. 2012 , Robinson & Wray 2012 ). Similarly, IUPC-radiotelemetry is a valuable in vivo method to study uterine contractile pressure throughout pregnancy and labor in non-anesthetized mice ( Pierce et al. 2010 , Rada et al. 2015 ). Transcervical IUPC placement in mice would provide a rapid, reliable, and quantifiable measure of contraction frequency, duration, strength, and resting uterine tone, using similar clinical methods as in humans ( Serr 1974 ). Here, we present a description for the placement of an IUPC through the cervical os of anesthetized pregnant mice for intermittent measurement of intrauterine contractile pressure. Our objectives were to: 1) optimize the experimental parameters including catheter size and anesthetic agent, 2) examine whether a transcervically-placed pressure catheter is sensitive enough to detect changes in IUP between different late-stage pregnancy days and intrapartum, 3) determine whether this method can be utilized numerous times in a single mouse pregnancy without affecting the timing of delivery and fetal outcome and 4) compare the in vivo contractile activity between mouse models of term and preterm labor.

The laboratory setup used to monitor IUP changes during mouse pregnancy is shown in Figure 1 . A pressure control unit (PCU-2000) served as a two-channel amplifier and interface between the pressure catheter and PowerLab data acquisition system ( Fig. 1A ). A speculum tube ( Fig. 1B ) was used to visualize the mouse cervix for IUPC placement. Figure 1C shows a close-up image of the pressure-sensitive transducer located at the tip of the catheter. In an initial subset of mice, postmortem visualization of the IUPC ( Fig. 1D ) allowed determination of the length needed to surpass the uterocervical junction for each gestational day examined. Optimal recordings of intrauterine pressure were achieved by adjusting the position of the catheter. Two sizes (1.4F and 3.5F) of pressure catheter were tested in order to determine which produced uterine contractile recordings that were similar to those previously reported using a surgically placed IUPC in mice. While there was no statistical difference in IUP measurements ( Fig. 2A ) obtained from the two different sized pressure catheters, the 1.4F catheter was more difficult to guide through the visible cervix, unlike that of the 3.5F catheter. Therefore, subsequent studies utilized only the 3.5F catheter. In order to minimize the effect of anesthetic agents on uterine tone, three different anesthetic agents were tested based on reports of smaller hemodynamic effects or prior use in intrauterine pressure studies: isoflurane, tribromoethanol (also known as Avertin) and urethane ( Mahendroo et al. 1999 , Roth et al. 2002 , Janssen et al. 2004 , Word et al. 2005 ). There was no statistically significant difference between IUP measurements obtained by using avertin, 1% isoflurane or urethane as the anesthetic agent. In contrast, 2% isoflurane significantly (p<0.05) inhibited IUP. While isoflurane is most commonly mixed with 100% O 2 for use as an anesthetic agent, the effect of hyperoxia on in vivo intrauterine pressure or contractility is unknown. Therefore, the anesthetic effect of 1% isoflurane mixed with 100% O 2 or normal air (21% O 2 , 78% N 2 and 0.4% CO 2 ) on IUP was compared. After 30-40 min exposure to 1% isoflurane mixed with room air, a decrease was observed in the respiratory rate and IUP, but not for mice anesthetized with 1% isoflurane mixed with 100% O 2 . By using a pulse oximeter, significant (p<0.05) differences were quantitatively detected in the heart rate, respiratory rate and oxygen saturation between the anesthetic agents ( Fig. 2B ). Based on the above results, Avertin or 1% isoflurane mixed with 100% O 2 were found to be optimal anesthetics for IUPC measurements. Next, we examined whether a transcervically-placed pressure catheter could detect differences in IUP between gestational day 16, 18, 19 and intrapartum, and whether mice could be utilized at multiple gestational time points to measure IUP, without negatively impacting IUP, timing of delivery, litter size or fetal outcome. Optimal intrauterine contractile pressure curves were rhythmic with an average contractile integral (AUC), amplitude, frequency, duration and Montevideo units per gestational day, as reported in Table 1 . The IUP recorded on days 16 and 18 of mouse pregnancy was low in amplitude, infrequent and were not significantly different from one another ( Fig. 3 ). However, IUP significantly increased on the morning (8-11 am) of day 19 and further increased during the afternoon (noon – 4pm) of day 19 of pregnancy, where it remained elevated until the intrapartum period (occurring between 7 pm on day 19 to 3 am on day 20). In post-hoc analysis, we identified an AUC threshold of 1.4 mmHg/sec where 88% of the measurements on the morning of day 19 were above the threshold (sensitivity) and 100% of the measurements on day 18 were below the threshold (specificity). Moreover, an AUC threshold of 2.77 mmHg/sec discriminates between in-active labor (the morning of day 19) and active labor (the afternoon of day 19 and intrapartum period) with 100% sensitivity and 94% specificity. The IUP was not significantly different between mice examined on a single experimental day compared to animals with measurements taken at all experimental time points (gestational days 16, 18 and 19). Moreover, measuring IUP at a single or multiple gestational time points did not negatively impact the increasing maternal weight (p>0.05) with advancing gestation, timing of delivery (p=0.91), offspring number (p=0.31) or fetal weight (p=0.44) ( Fig. 4 ). Following delivery of pups, there were no physical signs of uterine trauma due to insertion of the IUPC at multiple timepoints during pregnancy and active labor (data not shown). No dams were lost due to poor catheter placement, and no stillborn or resorbed fetuses were observed at the site of IUPC placement. Key differences between mouse and human pregnancy, which could impact uterine contractions during labor, include the number of offspring and uterine structure. Therefore, we examined the relationship between IUP and the number of offspring, intact uterine weight, uterine diameter and pup weight, as it changed during pregnancy ( Table 2 ). On average, the greatest coefficients of determination were between IUP (AUC, amplitude and Montevideo units) and uterine diameter (R 2 0.66 to 0.70) as well as between IUP and fetal weight (R 2 0.67 to 0.70). Conversely, offspring number and maternal weight was observed to correlate the least (R 2 0.07 to 0.14 and R 2 0.11 to 0.15, respectively) with IUP during pregnancy. In other words, at least 66% of the variation in IUP during pregnancy can be explained by the relationship to uterine diameter and fetal weight. Finally, IUP changes in two mouse models of PTL were monitored and compared to IUP changes observed during term pregnancy and labor. Given that measuring IUP at multiple gestational time points did not negatively affect IUP recordings ( Fig. 3 ) or pregnancy outcomes ( Fig. 4 ), intrauterine contractile pressure was measured at two time points prior to delivery, as well as intrapartum. Since LPS-treated mice are reported to deliver earlier than RU486-treated mice ( Migale et al. 2016 ), IUP was measured at different experimental time points for the two mouse models of PTL ( Fig. 5 ). By 4 hours post-injection, IUP was already significantly (p<0.05) greater in LPS-treated mice compared to saline-treated control mice. At 8 hours post-injection, IUP was significantly (p<0.05) greater than at 4 hours post-injection and remained elevated into the intrapartum period of LPS-treated mice. At 8 hours post-injection, RU486-treated mice had significantly (p<0.05) greater IUP compared to control mice. By 16 hours post-injection, the IUP of RU486-treated mice had doubled and remained elevated into the intrapartum period. LPS-treated and RU486-treated mice were monitored for their timing of delivery, which occurred at significantly (p<0.05) different time points (13.7 hours versus 23.6 hours post-injection, respectively), as well as significantly earlier than control-treated mice. The intrapartum IUP was compared between both mouse models of PTL (LPS- and RU486-induced) and normal term labor (gestation day 19.5), but no significant differences were observed in the AUC, amplitude, frequency or duration of contractions ( Table 3 ). Moreover, the Montevideo units calculated from the intrapartum period of mice undergoing term and preterm labor were similar.

Currently, an ex vivo isometric contractility assay using uterine tissue strips is the most widely-utilized method to evaluate uterine function in mouse models used to study parturition. However, there is a lack of non-invasive or minimally-invasive tools to rapidly assess in vivo uterine contractile activity. In the present study, we established a transcervical IUPC method that can successfully distinguish pre-labor (morning of day 19) and labor (afternoon of day 19 and intrapartum period) associated contractile changes in pregnant mice. Moreover, we found that this transcervical IUPC method can effectively monitor intrauterine contractile pressure at multiple gestational time points in a single mouse without compromising the timing of delivery or fetal outcomes. We established that at least 66% of the variation in IUP during mouse pregnancy can be explained by the relationship to uterine diameter and fetal weight. Additionally, we confirmed that IUP of experimentally-induced PTL was similar to that of normal, term labor of pregnant mice. Thus, transcervical IUPC approaches appear to be practical and very useful for monitoring IUP changes during the progression of labor in mice. The intrauterine contractile activities measured and reported in this study are in agreement to those from prior studies utilizing surgically implanted IUPCs in rats and mice. Intrauterine pressure was low-amplitude (3-5 mmHg) and irregular on day 16-18 of pregnancy ( Fuchs & Poblete 1970 , Buhimschi & Garfield 1996 ), but developed into regular contractions with increased intensity and duration on the day of parturition ( Deis & Pickford 1964 ). During the intrapartum period, the contraction duration and AUC reached maximal levels ( Shi et al. 2008 , Mackay et al. 2009 ). Similar to studies utilizing surgically implanted IUP-radiotelemeters in mice ( Pierce et al. 2010 , Rada et al. 2015 ), the non-surgical, transcervical IUPC approach described here was able to distinguish between pre-labor (day 19 morning) and labor (day 19 afternoon and intrapartum) contraction patterns in mice. The amplitude, and therefore AUC, of IUP changes recorded in the current study was greater than values previously reported from surgical-IUPC studies, but lower than surgically implanted IUPC-telemetry studies. One potential explanation includes the level of anesthesia required for surgical placement of the catheter and subsequent IUP recordings. Although Buhimschi and Garfield reported uterine activity that was similar between anesthetized and non-anesthetized rats, they noted large fluctuations (artifact signals) in their recording using non-anesthetized animals due to movement of rats ( Buhimschi & Garfield 1996 ). Thus, we decided to use short-term general anesthesia in our study. It is well established that injectable as well as inhaled anesthetics cause depressed cardiovascular and respiratory function. Specifically, pentobarbitol and a xylazine-ketamin mixture are reported to have the greatest hemodynamic effects, yet serve as the most commonly utilized anesthetic agents in animal studies, ( Furukawa et al. 1998 ). In our initial study, we chose to limit the anesthetics investigated to those with either a prior use in surgical-IUPC studies (avertin and urethane) or demonstrated to cause the least hemodynamic effects (inhaled isoflurane). We found that 2%-inhaled isoflurane was the only anesthetic agent to significantly inhibit IUP. However, urethane as well as 1% isoflurane mixed with room air had a significant dampening effect on the respiratory rate, while avertin use in room air had significantly diminished oxygen saturation. It appears that 1% isoflurane mixed with 100% oxygen is best suited for our repetitive IUP studies. Moreover, given the lack of pharmacological grade avertin and urethane, their use is becoming more restricted in research involving animals. There are advantages and disadvantages to utilizing a transcervical IUPC compared to other methods used to monitor in vivo uterine contractile activity of animal models for studying pregnancy and labor. The advantages include: non-surgical IUPC placement, ability to use at multiple experimental timepoints without negatively impacting pregnancy outcome or timing of delivery, precise as well as quantifiable recording of intrauterine contractile pressure activity, and the relatively inexpensive equipment costs. The disadvantages of using a non-surgical IUPC are the intermittent, non-continuous recording of in vivo uterine contractile activity throughout pregnancy, and the need to anesthetize mice. A primary weakness of our study is the inability to observe the exact intrauterine location of the pressure catheter during the recordings, which can be overcome in the future by utilizing ultrasound-guided catheter placement. The benefit of using a surgically-placed IUPC through the uterine wall is the ability to view the precise location where contractile recordings are being obtained. Furthermore, a specific benefit of utilizing a surgically-implanted IUPC-radiotelemeter is that it allows continuous recording throughput pregnancy in non-anesthetized mice. Another optimal non-surgical method to monitor IUP activity involves recording uterine electromyographic activity from the abdominal surface. This approach was developed by Garfield and colleagues using pregnant animals ( Buhimschi & Garfield 1996 , Buhimschi et al. 1998 ), and has been shown equally effective in women ( Freundlich & Wingate 1973 , Wolfs & van Leeuwen 1979 , Buhimschi et al. 1997 , Hadar et al. 2015 ). Unfortunately, this method has yet to be adapted for utilization in experimental mouse models to study pregnancy and parturition. In women, the amplitude of IUP increases from 25-30 mmHg to 60-65 mmHg during labor progression ( Buhimschi et al. 2003 ). In the current study, we found that the amplitude of IUP also increases from an average of 14 mmHg on the morning of the expected day of delivery to 20 mmHg during the pre-partum period. While the contractile amplitude of the pregnant mouse is slightly lower than women, the Montevideo units appear to be similar to women, at least 200 MVUs, as a result of greater contractile frequency observed in mice. Although mice are a polytocous species, the number of offspring was not found to be correlative with IUP during pregnancy. However, we found that at least 66% of the variation in IUP during mouse pregnancy can be explained by the relationship to uterine diameter and fetal weight, likely the result of “uterine stretch” as the stimulus for contractile activity. Gradual increase in intrauterine pressure leading to parturition in humans is well established. Moreover, several transgenic mouse models of altered parturition due to insufficient uterine activity substantiate the importance of uterine contractions required for successful parturition ( Ratajczak & Muglia 2008 ). However, little attention has been given to development of non-surgical methods for repeated and accurate measurement of IUP to assess labor onset in mice. The findings of this study established that non-surgical placement of an IUPC through the cervix could serve as an effective method to measure and monitor IUP changes during the progression of labor in anesthetized pregnant mice, while avoiding adverse pregnancy outcomes. The ability to repeatedly monitor IUP in mice prevents the need for tissue harvest at multiple gestational timepoints and promotes the principles of reduction and refinement in animal research. Future studies utilizing transcervical IUPC would also benefit pre-clinical mouse studies to determine the in vivo efficacy of novel tocolytic and uterotonic compounds on uterine contractile pressure ( Herington et al. 2015 ).

All experiments involving the use of mice had prior approval from the Institutional Animal Care and Use Committee at the Vanderbilt University Medical Center. Adult CD-1 wild-type (Charles River Laboratories, Raliegh, USA) mice were housed with free access to food and water in 12h light: 12h dark cycles. Virgin CD-1 (8-12 weeks, 25-30 grams) female mice were bred overnight with fertile males, and the presence of a copulatory vaginal plug was considered day 1 of pregnancy. Maternal weight was recorded on antepartum days 16, 18 and 19, and postpartum day 1. Timing of delivery was monitored using infrared video surveillance. The number of offspring, as well as their weights, was recorded on the morning of postnatal day 1. At necropsy, uteri were examined for signs of injury or fetal loss caused by placement of the pressure catheter, and the intact uterine weight was recorded as well as the uterine diameter by use of a caliper. For each of the experiments described below, time-point data were obtained from a minimum of 5 different animals. A progesterone receptor antagonist, mifepristone (also known as RU486; Sigma, St. Louis, USA) was dissolved in 30 μL 100% ETOH and then added to 200 μL sesame oil (Sigma) prior to subcutaneous injection (0.3 mg/230 μL per mouse) into the nape of the neck at 7 am on day 15 of pregnancy. Similarly, control animals received vehicle (30 μL 100% ETOH diluted 200 μL in sesame oil). Lipopolysaccharide (LPS; 100 μg; serotype O55:B5, Sigma) dissolved in 100 μL sterile normal saline or vehicle control saline (100 μL) was injected (i.p.) at 7 am on day 15 of pregnancy. On antepartum days 16, 18 and 19, as well as intrapartum, defined as the delivery of at least one pup, mice were anesthetized with either: 1% or 2% isoflurane (inhaled using a precision vaporizer and either 100% oxygen or room air), tribromoethanol (125-250 mg/kg, i.p.) or urethane (1-1.2 g/kg, i.p.; Sigma). Anesthetized mice were placed on a heating pad and then a speculum tube (0.5 cm width, 1.5 cm length) was inserted into the vaginal canal to visualize the cervix. A pressure catheter (Millar 1.4F and 3.5F, SPR-671 and SPR-524, respectively; ADInstruments, Colorado Springs, USA) was inserted transcervically to 3-6 cm depth, which was a predetermined length to pass the pressure sensing IUPC tip beyond the uterocervical junction. Mice were anesthetized and catheterized for each of the IUP recordings. After intrauterine contractions became rhythmic with consistent amplitude and frequency, baseline intrauterine contractile pressure was recorded (measured in mmHg) using PowerLab 8/35SP and PCU-2000 (ADInstruments) hardware. The IUPC was calibrated at baseline (0 mmHg) and 25 mmHg on each experimental day using the PCU-2000. The pressure catheter was cleaned with 1% tergazyme and then disinfected in CIDEX ® OPA per manufacturer’s instructions for 30min before each use in mice, as well as between mice. Intrauterine contractile pressure recordings were analyzed using LabChart 7 Pro software (ADInstruments). The contractile activity was assessed for a 10-minute period of time by: mean amplitude and frequency, AUC relative to baseline (normalized by dividing it by the exact number of seconds during the period), as well as Montevideo units (the sum of amplitudes during the 10-minute period of time). Respiratory rate (breaths per minute), heart rate (beats per minute) and oxygen saturation (%) was measured using a MouseOX Plus Pulse oximeter and physiological monitor (STARR Life Sciences Corp, Oakmont, USA). Six measurements were recorded per mouse at each experimental timepoint. For clarity, results are shown as mean ± SEM. Statistical analysis was performed using GraphPad Prism version 6.0e software (GraphPad Software, Inc., La Jolla, USA), and p<0.05 was considered significant. A two-way ANOVA was used for statistical comparison between gestational days 16, 18, 19 and intrapartum of mice used to measure IUP on a single timepoint versus every experimental timepoint. In mice measured over multiple days, we calculated all pairwise Pearson correlation coefficients and found the correlations were small (0.03 < r < 0.13). Conditional probability was calculated as the proportion of IUP above/below a threshold per gestational day or intrapartum period. The timing of delivery, offspring number and weight were analyzed using a one-way ANOVA to compare mice used to measure IUP on a single timepoint versus every experimental timepoint. Maternal weight was analyzed using a two-way ANOVA for statistical comparison between mice used for a single timepoint versus every experimental timepoint for IUP studies on gestational days 16, 18, 19 and postpartum day 1. To compare mouse models of PTL vs saline-treated control mice at multiple hours after treatment, a two-way ANOVA was performed. When ANOVA results were significant, pairwise tests were conducted using a Bonferroni adjustment for multiple comparisons. A Pearson correlation analysis was performed to compute the coefficient of determination (R 2 ).