Introduction

Preterm birth (PTB), defined as birth before 37 completed weeks of gestation, is a significant determinant of neonatal mortality, disease, and disability in surviving children [1]. Spontaneous premature labor (PTL) is the leading cause of PTB, accounting for approximately 45% of cases [1]. The risk of PTL is increased by prior preterm birth, preterm premature rupture of the membranes (PPROM), uterine overdistension, stress, and immunologically mediated processes [1, 2]. In attempts to prevent PTB, various tocolytics have been trialed for blocking spontaneous PTL, with the goal of extending pregnancy long enough for corticosteroid administration to mature the fetal lungs and improve birth outcomes. Tocolytic compounds examined include betamimetics, such as salbutamol and ritodrine, calcium ion (Ca2+) channel blockers (CCBs), such as nifedipine (NIF), Ca2+ competitors, such as magnesium sulfate, inhibitors of prostaglandin-endoperoxide synthase 2 (PTGS2), such as indomethacin (IND), oxytocin (OT) receptor antagonists (OTRA), such as Atosiban, and nitric oxide (NO) donors, such as nitroglycerine. These tocolytics are now well-characterized and work through either inhibiting pro-contraction signaling pathways, in particular by suppressing elevation of intracellular Ca2+ levels and blocking the availability and actions of uterotonins, or through activating pro-relaxation signaling pathways, in particular by raising intracellular cyclic adenosine monophosphate (cAMP) levels. Several systematic reviews and meta-analyses have been conducted in different parts of the world to assess the relative effectiveness of different tocolytic agents [3,4,5,6,7,8]. As a consequence, the guidelines for tocolytic management differ internationally [3, 9], and tocolytics licensed in one region of the world may or may not be licensed elsewhere. Where certain classes of tocolytics are not licensed, clinicians may utilize them off-label as second-line therapy. Approximately 75% of drugs used in obstetrics for tocolysis are unlicensed [10], this is because there are no specific therapeutic agents explicitly developed for tocolytic management except atosiban. Most tocolytics currently in use were developed for other medical indications, but were found to have tocolytic actions [10]. Moreover, there is no single agent currently available as a first-line tocolytic that is not associated with risks of side-effects [11]. Atosiban is associated with fewer side-effects than other tocolytics, but there is little evidence of efficacy [12]. All currently available tocolytic agents have relatively limited efficacy in postponing PTL and improving neonatal outcomes. As such, there remains a pressing need to evaluate the myometrial contraction blocking capabilities of novel drugs to improve tocolytic therapy.

Here, we report a comprehensive ex vivo analysis of three novel contraction-blocking agents: 2-aminoethoxydiphenyl borate (2-APB), glycyl-H-1152 dihydrochloride (GH), and HC-067047. We examined the potency of these “novel tocolytics” against spontaneous pregnant human myometrial contractions ex vivo and compared them to previously studied tocolytics, including NIF and IND, which have been used clinically for preventing PTB, as well as rolipram (ROL) and aminophylline (AMP), which have been previously investigated for uterine contraction inhibition.

2-APB was originally introduced as an inhibitor of inositol trisphosphate (IP3) receptors (IP3R) [13], which are a family of closely related Ca2+ channels embedded within the sarcoplasmic reticulum (SR). However, subsequent studies in intact cells have revealed that 2-APB also inhibits Ca2+ entry via store-operated channels (SOC); an effect that is independent of IP3R inhibition [14,15,16]. Thus, 2-APB has non-specific inhibitory effects on both IP3R and SOC, as well as on other Ca2+ transporters, e.g., sarcoplasmic Ca2+ ATPase (SRCA) pumps and transient receptor potential (TRP) channels of the TRPC family [17,18,19,20]. In maintaining Ca2+ homeostasis, G-protein coupled receptors, via activation of phospholipase C (PLC), generate intracellular IP3. IP3 then diffuses rapidly within sarcoplasm (cytoplasm) to bind with IP3Rs and release intracellular Ca2+ stores from the SR into the sarcoplasm [21,22,23]. The resulting increase in intracellular Ca2+ triggers the slow activation of SOCs in the sarcolemma (plasma membrane), which mediates store-operated Ca2+ entry (SOCE) into the sarcoplasm. Entry of Ca2+ through the sarcolemma contributes to the refilling of the SR Ca2+ stores and to the total intracellular pool of free Ca2+ in the sarcoplasm. Sarcoplasmic Ca2+ then binds to calmodulin, which activates the canonical signal transduction pathway that culminates in myosin light chain (MLC) phosphorylation and the actin-myosin cross bridge cycling that generates contractility [24, 25]. By inhibiting the activity of both SOC and IP3Rs, as well as other Ca2+ transporters, 2-APB prevents the elevation of sarcoplasmic Ca2+ levels, and in turn, inhibits myometrial contractility. Since its discovery [13], 2-APB has been tested on myometrial strips of rodents in several studies where it was found to inhibit both agonist-stimulated (OT, pennogenin tetraglycoside, Lannea acida plant extract, Ficus deltoidea plant extract) and spontaneous myometrial contractions [26,27,28,29,30,31,32]. However, despite potently inhibiting pregnant rodent uterine contractility, the effects of 2-APB on spontaneous pregnant human myometrial contractions were yet to be examined.

GH is a selective inhibitor of rho-kinase (ROCK). ROCK increases the sensitivity of uterine myocytes to Ca2+ through increased phosphorylation of MLC [33]. ROCK expression (mRNA abundance and protein levels) is upregulated at the end of pregnancy and is likely involved in the processes that underpin the increased myometrial contractility at term [34]. Thus, by inhibiting ROCK, GH reduces the sensitivity of uterine myocytes to Ca2+. GH has been found to inhibit both spontaneous and OT-stimulated pregnant human myometrial contractions ex vivo [35, 36]; however, comprehensive dose–response analyses were yet to be conducted to determine the potency of GH compared to other tocolytics.

HC-067047 is an inhibitor of transient receptor potential subfamily V, member 4 (TRPV4), a nonselective cation channel that is permeable to extracellular Ca2+ [37,38,39,40]. TRPV4 is activated by physiological stimuli, including stretch, swelling, heat, and pressure that may be relevant to human labor [41, 42]. Through inhibiting TRPV4 channels, HC-067047 prevents the influx of extracellular Ca2+ through these channels, thus preventing elevation of intracellular Ca2+ levels and myometrial contractility. TRPV4 is highly expressed in pregnant human myometrium and TRPV4 protein levels increase as gestation progresses [43]. These findings suggest that TRPV4 inhibition could be a potential novel tocolytic strategy [43, 44]; however, the effects of TRPV4 inhibitors, such as HC-067047, on spontaneous pregnant human myometrial contractions were yet to be investigated.

We also examined the non-selective phosphodiesterase (PDE) inhibitor, AMP, and the selective PDE4 inhibitor, ROL. By inhibiting PDEs, which are responsible for the breakdown of cAMP, both AMP and ROL raise intracellular cAMP levels, which promotes uterine relaxation. The tocolytic effects of theophylline, the active ingredient of AMP, and ROL have been reported in pregnant rodent and human myometrium [45,46,47,48,49,50,51]. However, their potency in suppressing spontaneous pregnant human myometrial contractions was yet to be comprehensively assessed. We also determined the potency of NIF and IND, which have well-documented tocolytic effects.

We conducted comprehensive dose–response analyses for 2-APB, GH, HC-067047, AMP, ROL, NIF, and IND using strips of pregnant human myometrium undergoing spontaneous contractions ex vivo. The location at which each of these agents affects uterine myocyte contraction signaling pathways is shown in Fig. 1. We then compared the contraction-blocking potency of the agents to assess the tocolytic potential of 2-APB, GH, and HC-067047 as novel tocolytics.

Fig. 1
figure 1

Overview of myometrial contraction signaling pathways and tocolytic action. The tocolytic agents examined block myometrial contractility through either inhibiting pro-contraction signaling pathways (2-APB, glycyl-H-1152, HC-067047, nifedipine, and indomethacin) or through activating pro-relaxation signaling pathways (aminophylline, rolipram)

Full size image

Materials and Methods

Drugs and Reagents

Drugs and reagents were obtained from the following sources: NIF (cat No N7634), IND (cat No I7378), and AMP (cat No A1755, MW-210.21) were purchased from Sigma-Aldrich Pty Ltd (Sydney, Australia); ROL (cat No 0905), GH (cat No 2485, batch-specific MW 458.3), 2-APB (cat No 1224), and HC-067047 (cat No 4100) were purchased from Tocris (Bristol, UK). Other miscellaneous reagents were purchased from Sigma-Aldrich Pty Ltd (Sydney, Australia) and ThermoFisher Scientific Inc (Sydney, Australia).

Human Myometrial Specimens

Biopsy specimens of human myometrial tissue were obtained from women, undergoing elective cesarean section at the John Hunter Hospital, NSW, Australia. Biopsies were collected with the approval of the Hunter and New England Area Human Research Ethics Committees (2019/ETH12330) and all participants gave informed written consent. Myometrial biopsies were obtained from the upper lip of the incision in the lower uterine segment. All myometrial biopsies were obtained from term pregnancies (37–40 weeks of gestation) where the woman was not-in-labor (NIL). The clinical indications for elective cesarean section were breech presentation or a previous cesarean. All women were examined clinically and those with signs of infection, or with diabetes mellitus, or treated with any medication other than prenatal vitamins were excluded from the study. Patient demographic data are shown in Table 1. Upon collection, the myometrial biopsies were placed in pre-chilled phosphate buffered saline (PBS) on ice for transportation and were used within 60 min to commence myometrial dose–response contraction assays.

Table 1 Demographic characteristics of the pregnant women
Full size table

Myometrial Contraction Assays

Myometrial contraction assays were performed as previously described [52,53,54] using an 8-channel Radnoti Tissue-Organ Bath System (Radnoti Glass Technology Inc., Monrovia, CA, USA) equipped with MLT0201 force transducers (ADInstruments, Bella Vista, NSW, Australia) and eight temperature-controlled organ baths. Human myometrial specimens were dissected into 8 × 1.5 × 1.5 mm tissue strips then connected to the force transducers using nylon thread and stainless-steel tissue clips (ADInstruments). Each strip was lowered into a separate organ bath containing 15 mL modified Krebs–Henseleit buffer solution (KREBS) (no Ca2+, no NaHCO3) (Sigma-Aldrich, cat no K3753-10X1L) supplemented to 2.5 mM CaCl2 and 25 mM NaHCO3. Organ baths were maintained at 37 °C and KREBS continuously gassed with 95% O2 and 5% CO2. The transducer position was adjusted to apply 1 g of tension to each strip. The strips were then equilibrated whereby every 10 min for a total of 30 min, the organ baths were drained then refilled with 15 mL of KREBS (tissue strips washed). Due to tissue creep during stabilization, tissue length was increased to return tension to 1 g. Washing and re-tensioning to 1 g was repeated twice more (each strip tensioned to 1 g a total of 3 times). Thereafter, tension stabilized between 0.5 and 0.9 g and strips were left to develop spontaneous rhythmic contractions ex vivo. Under the described conditions, the myometrial strips took approximately 2 h to establish spontaneous contractions with consistent amplitude and frequency. Testing protocols were then begun, and strips were then maintained under isometric conditions for the remainder of the experimental run. In control experiments, tension was maintained near baseline tension for 6–7 h (Fig. S1). In experiments involving serial drug additions, some treatments resulted in changes of baseline tension. To accommodate these baseline changes, AUC were calculated from the tension between contractions observed immediately prior to drug addition. Contraction data were captured and visualized in real-time using a PowerLab 8/35 data-acquisition system and LabChart software (ADInstruments). Contraction traces were analyzed for key contraction parameters, including amplitude (g) and frequency (contractions/h), and integration of these values to determine the area under the curve (AUC) (g tension × sec). AUC was considered an index of contraction performance and was calculated based on the total area for all contractions generated during each 30 min treatment window (Fig. S2).

Longevity of Spontaneous Contractions in Pregnant Human Myometrium Ex Vivo

To ensure that myometrial strips were able to contract for the duration of the tocolytic studies, strips were allowed to generate spontaneous contractions ex vivo for 7 h. Traces were then analyzed (in 60 min blocks) to confirm that there was no significant change in the resting tension, contraction amplitude, contraction frequency, or AUC across the 7 h period of spontaneous ex vivo contractions.

Dose–Response Study

Before administering drug treatments, a contraction baseline was established for each tissue strip during which 1 h of contractions of consistent amplitude and frequency was recorded. Following the establishment of the baseline, treatments were added to the organ baths and the effects on contractility recorded. For each tissue strip, cumulative concentrations of drugs were administered at 30 min intervals. The effect of each drug was assessed against each strip’s contraction baseline (each strip has an internal control). Seven drugs (2-APB, GH, AMP, HC-067047, ROL, NIF, IND) were analyzed against myometrium from different term pregnant NIL women (replicate numbers as indicated). To control for any effects of the drug vehicles (dimethyl sulfoxide (DMSO), Milli-Q water or KREBS buffer), equivalent cumulative volumes of vehicles were assessed against separate tissue strips during each contraction assay.

Data Analysis

Analysis of AUC was performed using LabChart 8.0 Pro with the dose–response module (ADInstruments). For each strip, the last 30 min of contractions immediately prior to commencing treatments was used as the baseline (100%). Effects of treatments were normalized against the baseline and data expressed as percent (%) of baseline contractility. Dose–response curves for AUC were generated using the non-linear regression model of GraphPad Prism 8.0 (GraphPad Software Inc., San Diego, CA, USA) and fitted through the (log inhibitor vs normalized response-variable slope) equation, Y = 100/(1 + 10^((Log IC50-X)*HillSlope)). The concentration of each drug required to inhibit ex vivo myometrial contractility by 50% (IC50) was determined as being a 50% reduction in total AUC relative to the contraction baseline. An ordinary one-way ANOVA followed by Dunnett’s multiple comparisons test was used to determine significant differences between the baseline and mean AUC of each dose used in the dose–response curve. A probability (P) value of < 0.05 was considered statistically significant.

Confirmation of IC50

We sought to confirm the accuracy of the IC50 values determined for each drug. Baseline contractility for tissue strips was recorded for 1 h. Each drug was then applied to individual contracting strips as a single treatment at the IC50 determined for each drug (with exception of HC-067047, which could not be solubilized at the required concentration predicted to be the IC50). Contractility was recorded for a further 1 h. The effect of administering each drug at the IC50 on AUC was then determined.

Results

Longevity of Spontaneous Contractions and Assessment of Drug Vehicles

We first sought to confirm that term NIL myometrial strips were able to maintain consistent spontaneous rhythmic contractions ex vivo for 7 h, which was the maximum duration of cumulative tocolytic treatments. The strips exhibited spontaneous rhythmic contractions within 2 h of the final equilibration wash/re-tension (Fig. S2, panel A). Contractions remained stable for over 7 h in that comparison of the 60 min periods revealed no significant changes in resting tension, contraction amplitude, frequency, or AUC (n = 5) across the 7 h period (Fig. S2, panels B–E).

During the treatment time courses (2.5–3.5 h), the administration of cumulative doses of DMSO (maximum of 0.42%), Milli-Q water, or KREBS (drug vehicles) had no effect on the contraction amplitude, frequency, or AUC (Fig. S3, panel A–C and Fig. 3, panel A–G).

Dose–Response Analyses

Each of the drugs analyzed dose-dependently inhibited contractions in spontaneously contracting strips of pregnant human myometrium ex vivo. The appropriate dosing regimens for each drug were optimized by prior organ bath contraction studies to determine a dose–response regimen for each drug, whereby the lowest drug concentration had no significant effect, and the highest concentration exerted the maximal inhibitory effect on contractions (ICmax). For all drugs except HC-067047, ICmax was the abolition of contractions (0% of baseline AUC) (data not shown). All drugs, except HC-067047, were therefore equally effective at abolishing spontaneous pregnant human myometrial contractions ex vivo but exhibited different potencies (different IC50 values) (Fig. 2).

Fig. 2
figure 2

Representative traces showing the effect of cumulative drug treatments on contractions. Spontaneously contracting pregnant human myometrial strips were treated with cumulative concentrations of A 2-APB (n = 8), B GH (n = 8), C HC-067047 (n = 7), D AMP (n = 8), E ROL (n = 8), F NIF (n = 10), and G IND (n = 10). Dotted lines indicate the points at which the treatment was added to the bath

Full size image

The cumulative concentrations of the novel tocolytics, 2-APB (1–120 µM), GH (0.1–80 µM), and HC-067047 (1–300 µM), affected spontaneous contractions similarly in that for each of these drugs, as contraction amplitude was reduced, contraction frequency significantly increased (Fig. 2, panels A–C). In contrast, as AMP (50–800 µM) and ROL (0.001–150 µM) inhibited contraction amplitude, contraction frequency significantly decreased (Fig. 2, panels D–E). This was particularly the case for ROL. The effects of NIF (0.1–50 nM) were consistent with the effects of 2-APB, GH, and HC-067047, in that as NIF reduced contraction amplitude, contraction frequency increased (Fig. 2, panel F), while the effects of IND (1–120 µM) were consistent with AMP and ROL, in that IND reduced contraction frequency as amplitude was inhibited (Fig. 2, panel G).

Tocolytic ICmax

2-APB and GH abolished spontaneous ex vivo contractions at 120 and 80 µM, respectively, whereas HC-067047 failed to completely abolish contractions even at the highest cumulative concentration tested (300 µM). At concentrations of ≥ 100 µM, HC-067047 precipitated out of solution within the organ baths. As such, the contractility recorded at the 100, 200, and 300 µM concentrations does not accurately reflect the effects of HC-067047 against spontaneous pregnant human myometrial contractions ex vivo. AMP and ROL abolished spontaneous contractions (ICmax) at 800 and 150 µM, respectively. The traditional tocolytics, NIF and IND, abolished contractions at 50 nM and 120 µM, respectively.

Determination of the Half-Maximal Inhibitory Concentration (IC50)

To determine tocolytic potency, dose–response curves were generated (Fig. 3, panels A–G) then the IC50 value determined for each drug. For the novel tocolytics, IC50 values were determined to be 53 µM for 2-APB (n = 8), 18.2 µM for GH (n = 8), and 48 µM for HC-067047 (n = 7). For the PDE inhibitors, IC50 values were 318.5 µM for AMP (n = 8) and 4.3 µM for ROL (n = 8), while for the traditional tocolytics, IC50 values were 10 nM for NIF (n = 10) and 59.5 µM for IND (n = 10). Contractility data are summarized in Table 2.

Fig. 3
figure 3

Dose–response curves. Plotted dose–response curves for the effect of A 2-APB (n = 8), B GH (n = 8), C HC-067047 (n = 7), D AMP (n = 8), E ROL (n = 8), F NIF (n = 10), and G IND (n = 10) on spontaneous pregnant human myometrial contractions ex vivo. Contractility was measured as AUC and expressed relative to the contraction baseline. Data are presented as mean ± SEM. There was a significant reduction in AUC in response to the cumulative doses of drugs. Comparisons were made between baseline and mean AUC of each dose using ordinary one-way ANOVA followed by Dunnett’s multiple comparisons test. A probability (P) value of < 0.05 was considered statistically significant. The asterisks indicate a significant difference, where 4 asterisks indicate P < 0.0001

Full size image
Table 2 Summary of experimentally determined IC50 and ICmax concentrations for each tocolytic
Full size table

Confirmation of Half-Maximal Inhibitory Concentration (IC50 )

Having determined the IC50 for each drug via the dose–response analyses, we then applied the IC50 to contracting strips as a single treatment to ascertain the accuracy of the determined IC50 values. Representative contraction traces are shown in Fig. 4, panels A–F. AUC analysis (1 h pre-treatment vs 1 h post-treatment) confirmed that for each drug, AUC was reduced by approximately 50% (2-APB = 49 ± 3% (n = 10); GH = 48 ± 3% (n = 10); AMP = 46 ± 4% (n = 10); ROL = 48 ± 5% (n = 5); NIF = 50 ± 3% (n = 10); IND = 49 ± 3% (n = 10) (Fig. 4, panel G).

Fig. 4
figure 4

Confirmation of the experimentally determined IC50 values. Representative traces showing the extent of contraction inhibition following treatment of pregnant human myometrial strips with the experimentally determined IC50 concentrations for A 2-APB (n = 10), B GH (n = 10), C AMP (n = 10), D ROL (n = 5), E NIF (n = 10), and F IND (n = 10). Black solid lines indicate the boundaries of the 1 h pre- and post-treatment analysis windows, and red dotted lines indicate the points at which the treatment was added to the bath. G The mean percent inhibition (relative to the contraction baselines) induced by each drug when applied at the IC50 for 60 min

Full size image

The Tocolytic Effect of the Drugs Is Reversible

To validate that the contraction inhibition was mediated by drugs and not due to diminished cellular viability and/or metabolic restriction of the tissue, the myometrial strips were washed with KREBS solution after tocolytic treatment to ascertain whether spontaneous contractions resumed. For all drugs analyzed, the strips resumed contracting after the washout procedure following a brief (< 30 min) recovery period (Fig. 5, panels A–G).

Fig. 5
figure 5

Reversibility of the tocolytics. Representative traces showing that after washout of the drugs, contractions spontaneously resumed in pregnant human myometrial strips treated with A 2-APB (n = 5), B GH (n = 5), C HC-067047 (n = 5), D AMP (n = 5), E ROL (n = 5), F NIF (n = 5), and G IND (n = 5). Red dotted lines indicate the points at which organ baths were drained then refilled with fresh KREBS buffer

Full size image

Discussion

Evidence suggests that PTL is a syndrome attributable to multiple pathological processes, including infection or inflammation, uterine overdistension, stress, ischemia or hemorrhage, endocrine disorders, immunologically mediated processes [55], and a gene expression pattern distinct from term labor [56]. Thus, PTL is a heterogeneous condition of multiple dysfunctions of preterm tissue that may lead to myometrial contractions, membrane/decidual activation, and/or cervical ripening. These three attributes constitute the common terminal pathway of both preterm and term birth. Hence, it is unclear whether spontaneous PTL results from premature activation of the term labor process, or due to pathological insults initiating uterine transformation from quiescence to overt labor [57]. Although evidence indicates that preterm and term labor are distinct processes [56], the preterm and term labor ultimately converge at the level of the myometrial contractile proteins. In this regard, there are no data available reporting that the levels of contractile proteins relevant to this study (IP3R, TRPV4, L-type Ca2+ channels, and ROCK) change between preterm and term human myometrium. Therefore, in the absence of such data, tocolytic agents that target these proteins should not be dismissed, as they may yet have relevance to tocolysis during PTL.

In this study, we performed comprehensive dose–response analyses to examine the contraction-blocking potency of three potential new tocolytics, 2-APB, GH, and HC-067047. We compared the IC50 of these new tocolytics to the PDE inhibitors, AMP and ROL, and to the clinically deployed tocolytics, NIF and IND. In terms of inhibiting spontaneous pregnant human myometrial contraction ex vivo (measured as AUC), the order of potency of the tocolytics from highest to lowest was NIF > ROL > GH > HC-067047 > 2-APB > IND > AMP (Table 3).

Table 3 Ranking of tocolytic agents according to their contraction blocking potency
Full size table

As a non-specific inhibitor of IP3Rs and SOC, 2-APB blocks the Ca2+ entry from both intracellular stores [13] and extracellular space [14,15,16], which prevents the elevation of cytosolic Ca2+ levels that drives contractions. Suppression of myometrial contractions in vitro by 2-APB was first reported by Ascher-Landsberg et al., who showed inhibition of both OT-stimulated and spontaneous contractions in rat myometrium, with contractions ultimately abolished at 100 µM [26]. This generally aligns with other subsequent studies in rodents, where 2-APB inhibited uterine contractions stimulated by different agonists, including OT and pennogenin tetraglycoside [26,27,28,29,30,31,32]. In our study, 2-APB abolished spontaneous pregnant human myometrial contractions at 120 µM, which is largely consistent with Ascher-Landsberg et al. [26], and we identified an IC50 (50% reduction in baseline AUC) for 2-APB of 53 µM. However, unlike Gravina et al., who reported a reduction in mouse myometrial contraction frequency following treatment [29], we observed that in pregnant human myometrium, contraction frequency increased as amplitude decreased in response to cumulative 2-APB treatments. Moreover, existing literature indicates that 2-APB exerts concentration-dependent biphasic effects on both SOC and IP3Rs. Patch-clamp studies on intact cell lines have shown that 2-APB has a stimulatory effect on Ca2+ entry via SOC and the IP3R gating system at lower concentrations (< 10 µM), and thus caused a transient increase in the amplitude of Ca2+ rise, whereas higher concentrations (> 10 µM) of 2-APB inhibits Ca2+ entry [15, 58, 59]. However, a biphasic effect of 2-APB was not observed in our myometrial tissue strip studies. We generated concentration–response curves based on assessment of both AUC (Fig. 3) and amplitude alone (data not shown) and in both cases, we detected no stimulatory effect of 2-APB on myometrial contractility at low concentrations (1 µM). The inconsistency may be attributable to the different experimental models in that the biphasic effect of 2-APB was observed against a non-uterine myocyte cell in monoculture using patch-clamp analysis, whereas our study utilized strips of pregnant human myometrium in a contraction bioassay system [15, 58, 59]. In support of this, existing literature demonstrates that isoform expression of STIM (1–2), a SR membrane protein that induces the opening of the SOCs, and ORAI (1–3), a plasma membrane protein that forms the pore of SOCs, differs between myometrial cells in culture and myometrial tissue [60,61,62]. 2-APB is shown to have differential effects on different STIM and ORAI isoforms [63, 64], which together mediate SOCE. Thus, the absence of a biphasic effect in the present study may be attributable to differences in expression of STIM and/or ORAI isoforms, compared to cell lines.

By inhibiting the isoenzyme, ROCK [65], GH reduces the Ca2+ sensitivity of uterine myocytes. The myometrial contraction and relaxation cycle depends on the equilibrium between phosphorylation and dephosphorylation of MLC, where myosin light chain kinase (MLCK) phosphorylates MLC to promote contraction and myosin light chain phosphatase (MLCP) dephosphorylates MLC to promote relaxation. MLCK is a Ca2+-dependent enzyme that is activated by the formation of the Ca2+-calmodulin complex in response to an intracellular Ca2+ surge [66,67,68,69]. In contrast, MLCP is negatively regulated by a Ca2+-independent mechanism where a regulatory subunit of MLCP is phosphorylated by ROCK, blocking the action MLCP and potentiating the effect of MLCK. Thus, the mechanism requires less Ca2+ to regulate MLC phosphorylation; a phenomenon called Ca2+ sensitization [33]. It has recently been demonstrated that ROCK-mediated Ca2+ sensitization of contractility can be induced by muscarinic or OT receptor stimulation in rat and human myometrium [70]. Hudson et al. and Aguilar et al. reported that cumulative concentrations and a single concentration (1 µM) of GH inhibited both OT-stimulated and spontaneous contraction of human myometrium ex vivo [35, 36]. Our comprehensive dose–response analyses add to these data as we have shown that in spontaneously contracting pregnant human myometrium ex vivo, the IC50 for GH is 18.2 µM and contractions are abolished at 80 µM. Our analyses also demonstrate that as GH inhibits contraction amplitude, contraction frequency increases, similar to 2-APB.

The third novel tocolytic we examined was HC-067047, a selective inhibitor of TRPV4 channels. TRPV4 plays a role in extracellular Ca2+ influx in response to various stimuli (stretch, swelling, heat, or pressure). There are limited studies examining the tocolytic effect of HC-067047. In mouse and rat uterine tissue strips, a single treatment with 1 µM HC-067047 inhibited contractions stimulated by OT and GSK1016790A [43, 71]. However, when used to treat pregnant human myometrium at the same concentration (1 µM), HC-067047 had only a slight inhibitory effect against OT-stimulated contractions [72].

Interestingly, Villegas et al. reported that activation of TRPV4 channels by the TRPV4 agonists, GSK1016790A and 4αPDD, resulted in the inhibition of OT-stimulated contractions in pregnant human myometrium [73]. This was attributed to an indirect effect whereby Ca2+ influx through TRPV4 channels causes K+ efflux via BKCa-channel activation, which, in turn, causes membrane hyperpolarization that inhibits L-type Ca2+ channels. This contrasts with the findings of Ying et al., who reported that GSK1016790A increased the contractility of mouse uterine tissue, which was then inhibited by HC-067047 [43]. Ying et al. also reported that HC-067047 inhibited OT-stimulated mouse uterine contractions, as well as delayed parturition in both RU486- and inflammation-induced mouse models of PTL [43]. Moreover, Singh et al. reported that HC-067047 inhibited GSK1016790A-induced contractility in murine strips ex vivo [71], while another ex vivo study with rat myometrial strips reported that the TRPV4 antagonist, RN1734, significantly decreased uterine contractility, whereas the TRPV4 agonist, RN1747, increased contractility [74]. The reason for the discrepancy in relation to the effects of the TRPV4 agonism is unclear but may be related to temporal and physical Ca2+ compartmentalization that plays a role in fine tuning contractility [72]. Nonetheless, our findings are consistent with prior studies from the mouse and rat, in that TRPV4 antagonism by HC-067047 unequivocally inhibited ex vivo spontaneous pregnant human myometrial contractions in a concentration-dependent manner (1, 10, 100 µM). Precipitation of HC-067047 at higher concentrations (200, 300 µM) means that the potency of HC-067047 is likely higher than that determined during this study. HC-067047 may therefore be a novel avenue for tocolysis; however, an effective delivery strategy may be required that overcomes the low aqueous solubility, such as delivery via uterine-targeted nanoliposomes [53] or vaginally administered mucus penetrating nanoparticles [75].

As a non-selective PDE inhibitor, AMP leads to the intracellular accumulation of cAMP, which operates through various mechanisms to promote uterine myocyte relaxation (see Fig. 1). Prior studies have demonstrated smooth muscle and myometrial contraction inhibition by AMP in rodents [48, 49, 76] and human [51, 77,78,79,80,81]. In an ex vivo study with pregnant human uterine strips, Bird et al. reported that AMP (40 and 100 µM) produced concentration-dependent inhibition of OT-stimulated contractions [77]. In another study, Verli et al. reported that increasing AMP concentrations (0.01 nM–10 µM) reduced OT-stimulated human myometrial contractions by 25% of baseline contractility [51]. Leroy et al., Berg et al., and Lai et al. demonstrated that AMP inhibited spontaneous pregnant human myometrial contractions in a concentration-dependent manner, where Leroy et al. and Berg et al. determined IC50 values for AMP of ≤ 100 µM [78, 80] and Lai et al. reported at 263 µM [81]. In the present study, AMP also inhibited spontaneous pregnant human myometrial contractions in a concentration-dependent manner; however, a high concentration (800 µM) of AMP was required to abolish contractility. Moreover, our determined IC50 for AMP of 318.5 µM is higher than that reported by Leroy et al., Berg et al. (100 µM), and Lai et al. (263 µM). The reason for these differences is currently unknown.

While AMP is a non-selective PDE inhibitor, ROL selectively inhibits PDE4, which is highly expressed in pregnant human myometrium at term [82]. Verli et al. reported that increasing concentrations of ROL (0.01 nM–10 µM) inhibited OT-stimulated contractions in pregnant human myometrial strips, and at the highest concentration tested of 10 µM, the authors found that ROL inhibited 62% of baseline contractility. Leroy et al., Bardou et al., and Martinez et al. reported that ROL inhibited spontaneous pregnant human myometrial contractions in a concentration-dependent manner, with 50% inhibition of contractility observed at 100 nM, 158 nM, and 22.7 µM, respectively [45, 78, 83]. In our analyses, ROL abolished ex vivo spontaneous pregnant human myometrial contractions at 150 µM and we determined the IC50 to be 4.3 µM. We also noted that for both AMP and ROL treatment, as contraction amplitude decreased, contraction frequency decreased; this contrasts the effects of 2-APB, GH, and HC-067047, where we observed that contraction frequency increased as amplitude was inhibited.

As an inhibitor of voltage-gated L-type Ca2+ channels, NIF blocks the influx of extracellular Ca2+ that underpins uterine myocyte contractility [84, 85]. The effects of NIF are well established [86, 87] and it is currently used clinically for tocolysis in various countries [88]. In the present study, we determined an IC50 for NIF of 10 nM. This is lower than the IC50 values reported by Moynihan et al. [87] (55 nM) and Ballejo et al. [89] (200 nM) against spontaneously contracting pregnant human myometrium, but is largely consistent with the IC50 values reported by Kuc et al. [90] (8.68 nM) and Longo et al. [91] (7.4 nM) against OT-stimulated contractions in pregnant human myometrium. Our determined IC50 of 10 nM, and with contractions abolished at 50 nM, makes NIF orders of magnitude more potent than all the other drugs examined in this study.

Within pregnant myometrium, IND blocks the synthesis of the prostaglandins that promote uterine contractions [92] and has been widely used as a tocolytic for many years. The tocolytic effect of IND was first reported by Vane et al. [93] and has been subsequently examined many times [94,95,96,97,98]. Across these studies, IND was reported to inhibit myometrial contractions at different concentrations and there is no emergent consensus as to an IC50. In our analyses, contractions were abolished by 120 µM IND and we determined the IND IC50 to be 59.5 µM. During prior studies on spontaneously contracting pregnant human myometrium, Arrowsmith et al. [94] and Johnson et al. [96] reported IND IC50 values of 35.4 and 278 µM, respectively, placing our findings within the range of these prior studies.

Strengths and Limitations

This study was the first to conduct a comprehensive dose–response study for novel tocolytics, 2-APB, GH, and HC-067047, to determine IC50 concentrations and compare their potency with clinically used tocolytics. A strength of this study was our confirmation of determined IC50 for each drug, where we confirmed that a ~ 50% reduction in baseline AUC was achieved (with exception of HC-067047 due to solubility limitations) when each drug was applied to contracting myometrial strips as a single treatment. These data support the accuracy of our determined IC50 values within our experimental setting and may provide insight into the relative tocolytic potency that could be expected from the different drugs in a clinical setting.

The study has only examined tocolytic potency against unstimulated (spontaneous) ex vivo contractions in term, NIL pregnant human myometrium. The authors acknowledge that there is increasing evidence that there are distinct differences between preterm and term labor [56], which may call into question the relevance of these data gleaned from term NIL myometrium. However, as previously mentioned, there are no data available indicating that levels of IP3R [99,100,101], TRPV4 [43], and ROCK [34, 102, 103] change between preterm and term pregnant human myometrium. Additionally, changes in expression of L-type Ca2+ channels [104,105,106,107], PTGS1 [108, 109], PTGS2 [109], and PDE4 [51, 110, 111] have been reported between preterm and term myometrium; however, each of these proteins was reported to exhibit higher expression in preterm myometrium than at term, suggesting that tocolytics targeting these proteins may actually have greater relevance during PTL than labor at term.

Lastly, this unstimulated model is standard technique for elucidating contraction pathways; however, further valuable insight would be garnered by examining agonist-stimulated contractions, as well as term IL myometrium and preterm NIL and IL myometrium.

Final Remarks

This study represents a comprehensive analysis of the myometrial contraction-blocking potency of the novel tocolytics, 2-APB, glycyl-H-1152, and HC-067047, and their potency comparison against the traditional tocolytics, nifedipine and indomethacin, as well as other potential candidates, rolipram and aminophylline (Table 2). Among the novel tocolytics, glycyl-H-1152 was the most potent followed by HC-067047 and 2-APB. Glycyl-H-1152 was also found to be a more potent inhibitor of ex vivo myometrial contractions than indomethacin and aminophylline, but less potent than nifedipine and rolipram, making glycyl-H-1152 the third most potent contraction blocker assessed (Table 3). These data provide us with greater insight into the contraction blocking potency of these drugs, with glycyl-H-1152 in particular emerging as a potential novel tocolytic due to its substantial potency. Glycyl-H-1152 may be an excellent candidate for encapsulation into uterine-targeted nanoliposomes [53] or vaginally administered mucus penetrating nanoparticles [75] as novel tocolytic strategies for preventing preterm birth. Such platforms may also facilitate the administration of hydrophobic drugs, such as HC-067047. Further studies are warranted to assess the tocolytic efficacy and safety of these agents in vivo using preterm birth models.