Lipopolysaccharide promotes contraction of uterine
myocytes via activation of Rho/ROCK
signaling pathways
James L. Hutchinson,1 Shalini P. Rajagopal, Mei Yuan, and Jane E. Norman
Medical Research Council Centre for Reproductive Health, University of Edinburgh,
The Queen’s Medical Research Institute, Edinburgh, UK
ABSTRACT Myometrial contraction is a central feature of labor. Although a link between infection and
preterm labor is widely accepted, surprisingly little is
known about the mechanisms coupling infection-induced inflammation to myocyte contractile machinery.
This study explores the myocyte response to pathogenderived ligands in vitro. The pregnant human myometrial cell line PHM1-41 and primary cultured uterine
myocytes responded to Toll-like receptor (TLR) ligands, including the bacterial wall component LPS,
which at 100 ng/ml increased contraction of cells
embedded within collagen gels over 72 h compared to
PBS. LPS-treated myocytes secreted inflammatory mediators, including prostaglandin F2, the cytokines
TNF- and IL-6, and a range of chemokines. The
contractile response to LPS required TLR4 signaling
and was independent of prostaglandin synthesis. Neutralizing TNF- had no effect on LPS-mediated contraction; however, the Rho-associated protein kinase
(ROCK) inhibitors Y-27632 (10 M) and GSK-269962
(50 nM) both abrogated the contractile response. The
finding of LPS-mediated contraction was supported by
a 1.38 0.072-fold (meanSE) increase in myosin
light-chain phosphorylation 48 h post-treatment, assessed by in-cell Western blot analysis. Together, these
data suggest that, in addition to modulating the local
inflammatory environment, pathogen-derived ligands
may directly promote myometrial contractility via Rho/
ROCK signaling, thus contributing to preterm labormediated preterm birth.—Hutchinson, J. L., Rajagopal,
S. P., Yuan, M., Norman, J. E. Lipopolysaccharide
promotes contraction of uterine myocytes via activation
of Rho/ROCK signaling pathways. FASEB J. 28,
000 – 000 (2014). www.fasebj.org
Key Words: infection labor myometrium preterm
There is an observed association between spontaneous preterm birth and infection in humans (1). Preterm
labor (PTL) can be reliably induced in animal models of
infection using either live bacteria or proinflammatory
Toll-like receptor (TLR) agonists delivered through various routes, including direct intrauterine injection (2).
Although proinflammatory mediators, such as prostaglandins, are well-established modulators of uterine contractility, the biochemical mechanisms linking infection/
inflammation and myometrial contraction are unknown.
The TLR4 ligand lipopolysaccharide (LPS) induces
inflammatory responses, including proinflammatory cytokine release in myometrium in culture (3), and
interleukin (IL)-1 itself has been shown to facilitate
ligand-induced basal and store-operated calcium entry,
thus potentiating myocyte contraction (4). In addition,
LPS induces production of known contractile effectors,
such as prostaglandins in a broad range of cell types
(5). Further evidence for the role of TLR activation in
PTL is shown by the link between TLR4 polymorphisms
and preterm birth in humans (6), while administration
of TLR4 antagonists, or inhibitors of the major downstream target nuclear factor B (NF-B), delays the
onset of PTL in animal models (reviewed in ref. 7).
However, despite the widespread use of LPS and other
TLR agonists in vitro and in vivo to simulate inflammation and PTL, there are no direct data on the ability of
these agonists themselves to stimulate myometrial con-
1 Correspondence: MRC Centre for Reproductive Health,
University of Edinburgh, The Queen’s Medical Research
Institute, 47 Little France Crescent, Edinburgh EH16 4TJ,
UK. E-mail: [email protected]
doi: 10.1096/fj.13-237040
Abbreviations: CM, conditioned medium; DMEM, Dulbecco’s modified Eagle’s medium; ELISA, enzyme-linked immunosorbent assay; FCS, fetal calf serum; FGF, fibroblast growth
factor; GAPDH, glyceraldehyde-3-phosphate dehydrogenase;
G-CSF, granulocyte colony-stimulating factor; GM-CSF, granulocyte macrophage colony-stimulating factor; IFN, interferon ; IL, interleukin; LBP, LPS-binding protein; LPS,
lipopolysaccharide; MCP-1, monocyte chemotactic protein 1;
MIP, macrophage inflammatory protein; MKL1, megakaryoblastic leukemia 1; MLCP, myosin light-chain phosphatase;
MYL20, myosin light chain; NF-B, nuclear factor B; PBS,
phosphate-buffered saline; PGF2, prostaglandin F2; PHM1-
41, pregnant human myometrial 1-41; PTGS2, prostaglandin
synthase 2; PTL, preterm labor; qPCR, quantitative polymerase chain reaction; ROCK, Rho-associated protein kinase;
RT-PCR, reverse transcription polymerase chain reaction;
TLR, Toll-like receptor; TNF-, tumor necrosis factor ;
UtSMC, uterine smooth muscle cell; VEGF, vascular endothelial growth factor
0892-6638/14/0028-0001 © FASEB 1
The FASEB Journal article fj.13-237040. Published online September 27, 2013.
tractility. The studies described here address this issue
by exploring the effects of TLR agonists, particularly
LPS, on the contractile apparatus of the uterine myocyte.
MATERIALS AND METHODS
Myometrial biopsies were obtained through the Edinburgh
Reproductive Tissues BioBank from 11 currently healthy
women with a mean sem age of 34.4 3.6 yr, undergoing
elective Caesarean section at term (38–42 wk gestation) prior
to the onset of labor. Patients with medical complications,
such as diabetes, preeclampsia, hypertension, pyrexia, or
other clinical signs of infection, were excluded. Tissues were
taken postdelivery from the upper margin of the transverse
lower uterine incision, as described previously (8). All tissue
samples were collected from participant tissue donors with
informed and written consent, according to the ethical approval and governance granted to the Edinburgh Reproductive Tissues BioBank by the West of Scotland Research Ethics
Committee 4 (09/S0704/3).
Pregnant human myometrial 1-41 (PHM1-41) cells were a
kind gift from B. Sanborn (Colorado State University, Fort
Collins, CO, USA). Primary uterine smooth muscle cells
(UtSMCs; prepared by enzymatic dispersion from a nonpregnant donor) were obtained from Lonza (Slough, UK). Rat tail
collagen type I, prepared according to Bell et al. (9), was
obtained from A. Henke (University Clinic Cologne, Cologne, Germany).
The TLR4 antagonist LPS-RS and all TLR agonists were
purchased from Life Technologies (Paisley, UK), with the
exception of TLR4-grade LPS (Escherichia coli, EH100, Ra)
from Enzo Life Sciences (Exeter, UK). Other reagents used
were CCG-1423 from Enzo Life Sciences, GSK-269962 from
Tocris (Bristol, UK), oxytocin from Alliance Pharmaceuticals
(Chippenham, UK), AL-8810, Y-27632, and nifedipine from
Sigma (Poole, UK), tumor necrosis factor (TNF-) from
Peprotech (London, UK), -MYL20, -pMYL20, and -TNF-
antibodies from Cell Signaling Technology (Hitchin, UK),
donkey -mouse-680 and -rabbit-800 antibodies from Li-Cor
Biotechnology (Cambridge, UK), and rabbit -glyceraldehyde-3-phosphate dehydrogenase (-GAPDH) from Sigma.
Reagents were dissolved in phosphate-buffered saline (PBS)
solvent with the exceptions of indomethacin, Y-27632, and
GSK-269962, which used ethanol and nifedipine, and CCG-
1423, which used DMSO. Concentration-matched vehicle
controls were employed as indicated.
Tissue culture
PHM1-41 cells and primary UtSMCs were cultured in highglucose Dulbecco’s modified Eagle’s medium (DMEM;
Lonza) supplemented with l-glutamine (Sigma), 10% fetal
calf serum (FCS) and penicillin/streptomycin (both PAA
Laboratories, Yeovil, UK). PHM1-41 medium for routine
culture was additionally supplemented with 0.1 mg/ml G418
(Merck Chemicals, Nottingham, UK) and 0.1 mg/ml normocin (Life Technologies), which were removed one passage
prior to performing experiments. TLR ligand treatments, LPS
time-course treatments, and generation of conditioned medium (CM) for contraction assays were performed in triplicate using 6-well plates, 2 105 cells/dish, using 2 ml
medium supplemented with 5% FCS. Conditioned culture
supernatants were used immediately or stored at 80°C for
later analysis.
Gel contraction assays
Collagen gels for contraction assays were prepared in 24-well
plates at 105 cells/well as described by Dallot et al. (10).
Following overnight equilibration, gels were detached and
incubated with medium conditioned for 24 h with PHM1-41
cells treated in 6-well plates as indicated in Results. Inhibitor
concentrations were chosen on the basis of reported IC50
values. Gel treatments were performed in quadruplicate or
greater. Gels were visualized and photographed using a Leica
MZ6 microscope with Firecam camera system (Leica Microsystems, Wetzlar, Germany). Gel area was calculated using
Adobe Photoshop CS3 (Adobe Systems, San Jose, CA, USA).
Viability assays
Viable cell numbers were determined using the CellTiter 96
AQueous One assay (Promega, Southampton, UK). Gels or
monolayers were washed in PBS, then serum-free DMEM,
before resuspending in fresh medium plus 20% Aqueous One
reagent and incubation for 2 h under standard culture
conditions. Medium was transferred to 96-well plates, and
absorption was read at 490 nm.
Protein analysis and real-time reverse transcription
polymerase chain reaction (RT-PCR)
IL-6 and IL-8 Duoset enzyme-linked immunosorbent assays
(ELISAs) and Fluorokine MAP assays (R&D Systems, Abingdon, UK) were performed according to manufacturer’s instructions. RNA was extracted using Qiazol separation followed by RNeasy column clean-up with on-column DNase
digest (Qiagen, Manchester, UK) according to manufacturer’s protocols. cDNA was prepared using SuperScript VILO
(Life Technologies). TaqMan quantitative PCR (qPCR) for
prostaglandin synthase 2 (PTGS2) was performed using the
following primer/probe sequences: forward CCTTCCTCCTGTGCCTGATG, reverse: ACAATCTCATTTGAATCAGGAAGCT, and probe: TGCCCGACTCCCTTGGGTGTCA. Statistical analysis of qPCR data was performed using Ct values for
genes of interest, adjusted according to the geometric mean
average of those of -actin and 18S (Applied Biosystems).
In-cell Western blot analyses
In-cell Western blot analyses for total and phosphorylated
myosin light chain (MYL20) were performed as described by
Aguilar et al. (11), except that serum was retained throughout. Briefly, treatments were performed on cells seeded into
96-well plates and incubated overnight in 200 l medium plus
5% FCS. Ligands or PBS vehicle were diluted in (serum-free)
medium and added using a multichannel pipette to a final
volume of 10%, with treatments alternating by row. Samples
were fixed by the addition of concentrated formalin to 10%.
Following primary and secondary antibody incubations,
plates were analyzed using a Li-Cor Odyssey scanner (Li-Cor
Biosciences, Lincoln, NE, USA).
Organ bath
Freshly obtained longitudinal myometrial tissue strips were
dissected into 2- 2- 15-mm lengths and mounted in
organ tissue baths under isometric conditions with a resting
tension of 20 mN, as described previously (12). Myometrial
strips were allowed to equilibrate for at least 2 h in 10 ml
Krebs-Henseleit solution at 37°C (pH 7.4) with washing.
Following establishment of spontaneous contractions, strips
2 Vol. 28 January 2014 The FASEB Journal www.fasebj.org HUTCHINSON ET AL.
were treated with 1 ng/ml to 1 g/ml LPS alongside patientand time-matched control strips treated with PBS vehicle
alone. Contractility was recorded using an MLT0201 tension
transducer attached to one end of each strip and linked to a
PowerLab 8/30 data acquisition system running LabChart 6
(AD Instruments, Hastings, UK). Strips were treated at approximately half-hour intervals. For each treatment, 3 contractions were recorded. Averaged contraction rate, peak
amplitude (maximum-minimum), and force integral (area
under curve) were calculated as a percentage of pretreatment
values for each strip for analysis. Samples were further
normalized for graphical presentation as a percentage of
time-matched vehicle average for each LPS dose assayed.
Statistics
Results were analyzed by 1- or 2-way ANOVA, or t test,
following log transformation to normalize distributions where
necessary. Matched sample tests were employed where appropriate to control for batch effects. For ANOVA, comparison
was made between selected sample pairings using Bonferroni
post hoc tests. Where mean values are reported, the standard
error is indicated. All experiments were performed multiple
times as indicated in the figure legends. Where absolute
values are reported, figures display the results of 3 independent experiments. Where results are reported relative to
control values, data from a representative experiment are
shown.
RESULTS
Uterine myocytes respond to TLR agonists
We first set out to determine whether PHM1-41 cells
were capable of responding to TLR agonists, using the
immortalized PHM1-41 cell line (13). Following 24 h
treatment with agonists to each of TLRs 1–8, PHM1-41
supernatants were assayed for levels of the labor-associated cytokines IL-6 and IL-8. Levels of both cytokines
were significantly up-regulated in response to ligands
for TLRs 1 and 4–6 (Fig. 1A, B). The magnitude of
response seen was greatest in both cases for ligands to
TLRs 4 (LPS) and 5 (flagellin). We next chose to look
at the response to LPS in greater detail. Changes in
PTGS2 levels are closely associated with the onset of
myometrial contraction (14). PTGS2 mRNA levels in
PHM1-41 cells treated with LPS across a time course of
6–72 h showed a biphasic response by qPCR, that was
significantly elevated in comparison to vehicle at all
time points (Fig. 1C). Since PTGS2 was elevated by LPS
treatment, and prostaglandin F2 (PGF2) is known to
have profound stimulator effects on myocyte contractions, we next examined levels of PGF2 in the
PHM1-41 time-course supernatants. PGF2 was detected at a low level in vehicle-treated samples but was
significantly elevated after treatment with TLR4 agonist, levels rising steadily with time (Fig. 1D).
To further characterize the myocyte response to
bacterial endotoxin, we used the fluorokine MAP multiplex bead array system to analyze simultaneously the
production of 20 cytokines following PHM1-41 treatment with vehicle or the TLR4 ligand LPS, across a time
course of 6–72 h. Of these 20 cytokines, 9 were
undetectable for either treatment at any time point.
These were interferon (IFN;
Figure 1. A, B) PHM1-41 secretion of IL-6 (A) and IL-8 (B) in response to
stimulation with TLR agonists. Cells were exposed to agonists of each of the
TLRs 1–8, and supernatants were harvested for ELISA after 24 h. n 4
independent experiments. C) PHM1-41 PTGS2 levels in response to LPS
treatment over time. Cells exposed to LPS (squares with dashed lines) or PBS
vehicle (circles with solid lines) for between 6 and 72 h in 3 independent
experiments were harvested and analyzed by qPCR. D) PHM1-41 PGF2
secretion in response to LPS. Supernatants from C were analyzed by ELISA for
PGF2 for each time point. *P
LPS MODULATION OF UTERINE MYOCYTE CONTRACTION 3
pg/ml), IL-17 (
1.10 pg/ml), and macrophage inflammatory protein (MIP)-1/CCL4 (
0.72 pg/ml); values
in parentheses indicate manufacturer’s reported assay
sensitivity. Of the 11 detected cytokines, all showed
significant up-regulation in response to LPS (Fig. 2). Of
these, TNF- peaked at 6 h (Fig. 2J), IL-8/CXCL8 levels
peaked at 24 h (Fig. 2G), and monocyte chemotactic
protein 1 (MCP-1)/CCL2 concentration peaked at 48 h
following LPS treatment (Fig. 2H). The remaining
cytokines, CCL-5, CXCL-5/ENA-78, fibroblast growth
factor (FGF) basic, granulocyte colony-stimulating factor (G-CSF), granulocyte macrophage colony-stimulating factor (GM-CSF), IL-6, MIP-1/CCL3, and vascular
endothelial growth factor (VEGF), showed a progressive increase in output following LPS treatment over
the duration of the experiment.
LPS treatment induces a delayed contractile response
in uterine myocytes
Elevated myometrial IL-6 and, especially, PGF2 are
temporally associated with uterine contraction and
labor, while TNF- stimulates production of contraction-associated proteins, and its receptor levels are
CCL5/RANTES
J K Figure 2. PHM1-41 cytokine secretion
in response to LPS treatment over
time. Supernatants from cells treated
with LPS or PBS vehicle for between 6
and 72 h in 3 independent experiments were subject to Fluorokine MAP
analysis for the presence of 20 cytokines; 11 cytokines were expressed at
detectable levels: CCL5/RANTES (A),
CXCL5/ENA-78 (B), FGF basic (C),
G-CSF (D), GM-CSF (E), IL-6 (F), IL-
8/CXCL8 (G), MCP-1/CCL2 (H),
MIP-1/CCL3 (I), TNF- (J), and VEGF (K). Circles with solid lines denote values for vehicle-treated samples; squares with
dashed lines denote those for LPS. Time points at which levels were significantly greater in the LPS compared to the
vehicle-treated group are indicated: *P
4 Vol. 28 January 2014 The FASEB Journal www.fasebj.org HUTCHINSON ET AL.
increased in laboring myometrium (14–16). Given that
LPS modulated secretion of these factors, we therefore
chose to examine whether LPS stimulates myocyte
contraction directly, using both ex vivo tissue studies
and in vitro collagen gel contraction. Using freshly
obtained myometrial strips from term pregnant
women, we first examined short-term (
6 h), spontaneous ex vivo contractile behavior in response to treatment with 1 ng/ml to 1 g/ml LPS. Neither frequency
(Fig. 3A), nor area under the curve (Fig. 3B), nor peak
amplitude (Fig. 3C), was significantly altered in comparison to appropriate time-matched vehicle.
To supplement the organ bath data and extend to
later time points, we chose to use an in vitro system,
which allows monitoring of myocyte contraction in
isolation from other cell types and over a longer period
of time. The collagen gel contraction assay is an established procedure, whereby uterine myocytes are seeded
into a collagen gel lattice, whose contraction over time
is then monitored macroscopically (10). Following 6 h
treatment with 100 ng/ml LPS, no net change in
average gel size was observed relative to vehicle control
gels (LPS average gel area 98.81.62%, vehicle,
1001.94%, ns), whereas both 10 nM oxytocin and 10
ng/ml TNF- modestly but significantly enhanced basal
contraction over this time period (evidenced as a
reduction in gel area; Fig. 3D) with gel areas averaging
94.0 1.10% (P
0.001) and 96.6 1.42% (P
0.01) of
vehicle, respectively. Myocyte contraction is ultimately
mediated by myosin, whose activity is determined by
phosphorylation of serine residues on the regulatory
light-chain MYL20. Changes in MYL20 phosphorylation
following acute (
3 min) oxytocin treatment are readily detected using a sensitive and specific high-throughput in-cell Western assay format (11). While we were
able to confirm oxytocin-mediated induction of MYL20
phosphorylation using this approach, levels following
acute LPS treatment were similar to those with vehicle
(Fig. 3E), in line with the lack of effect of LPS in gel
contraction assays at 6 h and organ bath experiments.
The inflammatory process is multistaged, and LPS
signals via multiple pathways, each with differing kinetics, causing gene-level changes and inducing production of various inflammatory mediators capable of
acting in an autocrine fashion (17, 18). PTL following
LPS treatment in mice can take several days to develop
(19), while chronic infection is postulated in cases of
PTL in humans (20). In our PHM1-41 cell culture, we
observed that levels of many secreted cytokines continued to rise progressively well beyond the 6-h time
period studied in our organ bath and initial cell-based
contraction assay. Thus, we reasoned that any effects of
LPS (compared with vehicle) on contraction might not
become apparent until after the 6-h time period of
initial study. Measurement of gel area at 24-h intervals
up to 96 h following exposure to CM revealed that both
LPS and TNF- treatment resulted in significantly
reduced gel size at 48 and 72 h vs. vehicle (and 96 h for
TNF-; Fig. 4A–D). Analysis of the relative contraction
rate (rather than gel area, which represents total accumulated contraction) suggested that contractile enhancement compared to vehicle was maximal around
Figure 3. A–C) Average contraction
frequency (A), average force (B),
and average normalized peak amplitude (C) of isolated myometrial
strips treated with 1–1000 ng/ml
LPS as a percentage of vehicle
(Veh). n 11 patients. D) PHM1-41
collagen gel contraction in response
to short-term oxytocin (OT), LPS, or
TNF- stimulation. Area of collagen
gels embedded with PHM1-41 cells
was assessed following 6 h treatment
with OT (10 nM), LPS (100 ng/ml),
or TNF- (10 ng/ml) and compared to that seen for PBS vehicle treatment. n 6 replicates/treatment, representative
figure from 1 of 3 independent experiments. E) MYL20 levels following acute PHM1-41 stimulation with OT or LPS. Cells
seeded in 96-well plates were exposed to PBS, LPS, or OT for 20–180 s before rapid fixation and in-cell Western blot analysis
for phosphorylated MYL20. Data are expressed as normalized pMYL20 vs. time point vehicle, corrected against GAPDH.
n 6–8 replicates/treatment/time point, representative figure from 1 of 2 independent experiments. ns, not significant.
0.05, ***P
0.001 vs. vehicle.
LPS MODULATION OF UTERINE MYOCYTE CONTRACTION 5
48 h (Fig. 4E). Collagen gel contraction increases with
cell number (9). Notably, however, the observed difference between vehicle- and LPS-treated samples in our
study was not accounted for by LPS-stimulated proliferation, as LPS treatment caused a slight reduction in
viable cell numbers compared to vehicle (Fig. 4F). A
similar contractile response was observed following
substitution of PHM1-41 with primary cultured uterine
smooth muscle cells (Fig. 4G), suggesting that contraction in response to LPS- and TNF- is a general feature
of myometrial myocytes in vitro.
LPS-mediated contraction is TLR4 dependent and
TNF- independent
Endotoxin contamination is a known issue with LPS
preparations, causing responses independent of TLR4
(21). Furthermore, depending on bacterial strain and
preparation integrity, LPS can activate cells via alternative CD14- and LPS-binding protein (LBP)-independent pathways (22). To test the specificity of the TLR4
response, we repeated our experiments after 30-min
preincubation with LPS-RS, a TLR4 antagonist. Preincubation with LPS-RS dose-dependently inhibited IL-6
production by 24 h in PHM1-41 cells following LPS
treatment (Fig. 5A). In addition, LPS induced PHM1-41
cell IL-6 production was substantially lower in the
absence of serum, implying a requirement for LBP in
the response to LPS. Notably, LPS-RS at 1 g/ml (a
dose sufficient to abrogate LPS-induced cytokine production) also inhibited LPS (100 ng/ml)-induced contraction (Fig. 5B), indicating that LPS enhances uterine
myocyte contraction via TLR4-dependent pathways.
TNF- is an important downstream effector of LPS
signaling and contributes to LPS induction of PTL in a
mouse model (23). Since LPS treatment of PHM1-41
cells induces secretion of TNF-, which was itself procontractile in our assays on exogenous addition, we
next performed blocking antibody experiments to determine whether endogenously produced TNF- was
responsible for the observed LPS effects on contraction. Pretreatment of conditioned media with 50 ng/ml
TNF-blocking antibody was unable to prevent LPSFigure 4. A–D) PHM1-41 collagen gel
contraction in response to extended
LPS or TNF- exposure. CM from
PHM1-41 cells treated with PBS, LPS,
or TNF- was incubated with fresh
PHM1-41 cells embedded in collagen
gels, with gel area monitored and
recorded at 24 h (A), 48 h (B), 72 h
(C), and 96 h (D). E) Data from A–D,
analyzed to show a normalized rate of
contraction for each 24-h time period, expressed as the reciprocal of
the relative sample gel area, itself calculated as a fraction of the same gel area at previous time point. n 6 replicates/treatment, representative figure from 1
of 3 independent experiments. F) PHM1-41 cell viability in CM-treated collagen gels. Following final area measurement,
PHM1-41 gel viable cell number was compared between treatment groups using a colorimetric assay. G) Contraction of
collagen gels embedded with cultured UtSMCs at 72 h. n 6 replicates/treatment, representative figure from 1 of 3
independent experiments. *P
0.0001 vs. vehicle (Veh).
6 Vol. 28 January 2014 The FASEB Journal www.fasebj.org HUTCHINSON ET AL.
mediated contraction (Fig. 5C), despite fully inhibiting
IL-6 and IL-8 release stimulated by 1 ng/ml exogenous
TNF- (Fig. 5D, E). Thus, the procontractile effects of
LPS are independent of TNF- production.
LPS modulates myocyte contractility via
prostaglandin-dependent and independent pathways
Increased production of PGF2 following LPS treatment
could be a mechanism by which LPS induces myocyte
contraction. Indomethacin treatment at 50 M signifi-
cantly attenuated basal contraction of PHM1-41 collagen
gels (P
0.0001) but had no apparent effect on the
LPS-mediated component, given that LPS-treated mean
gel area at 72 h in both the presence and absence of
indomethacin remained at a similar percentage of
relevant vehicle (90.13.14 vs. 1002.46 and 90.0
3.68 vs. 1002.93% respectively, both P
0.05; Fig. 6A).
Cell viability remained unaffected by indomethacin
treatment at 72 h (Fig. 6B), with PGF2 levels remaining below the threshold of detection in indomethacintreated culture supernatants (not shown). Since indomethacin targets prostaglandin synthases, it has a
global inhibitory effect on prostaglandin production.
To look more specifically at the role of PGF2, we
performed a time-ourse experiment using the FP receptor antagonist AL8810. Two-way analysis of the effect of
vehicle- vs. LPS-treated cells at 24 h with and without
AL8810 revealed a significant interaction effect
0.05). AL8810 cotreatment with LPS-CM caused a
reduction in gel contraction, i.e., increased gel size,
compared to AL8810-treated vehicle at this time point
0.01; Fig. 6C). In contrast, in the absence of
AL8810, no net alteration in contractility of LPS-treated
gels was observed at this time point, in accordance with
previous experiments. At 48 h, however, this inhibitory
effect of LPS on net contraction of AL8810-treated gels
was no longer evident (Fig. 6D), and this loss of
inhibition was preserved at 72 h (Fig. 6E). Instead,
contraction rate analysis showed that a significantly
enhanced contraction was observed with LPS treatment
at 48 h compared to vehicle in the presence of AL8810
0.001; Fig. 6F). AL8810 had no effect on cell
viability (Fig. 6G). Together, these results suggest that
LPS stimulates production of a variety of prostaglandin
species that modulate contraction in a functionally
antagonistic manner, but that the LPS-stimulated contraction observed at later time points is independent of
prostaglandin activity.
LPS promotes myocyte contraction via increased
Rho-associated protein kinase (ROCK) activity
L-type calcium channels are the major species responsible for extracellular calcium influx in uterine myocytes (24). Calcium influx activates the calcium sensor
calmodulin, which binds to myosin light chain kinase,
activating MYL20 phosphorylation and subsequent contraction. L-type calcium channel activity is inhibited by
the tocolytic agent nifedipine. LPS was able to induce
contraction at 72 h, even in the presence of 20 g/ml
nifedipine (Fig. 7A), suggesting that its action is mediated by mechanisms other than L-type channel-mediated calcium influx, although the adverse effects of
nifedipine on cell viability means these data should be
interpreted with caution (Fig. 7B).
Figure 5. A) Effect of TLR4 competitive antagonism and serum omission
on PHM1-41 LPS-stimulated IL-6 secretion. B) PHM1-41 collagen gel
contraction at 72 h in response to
LPS, with and without LPS-RS. C)
Effect of TNF- neutralization on
LPS-induced PHM1-41 collagen gel
contraction at 72 h. n 6 replicates/
treatment, representative figure from
1 of 2 independent experiments. D,
E) Secreted IL-6 (D) and IL-8 (E)
levels following PHM1-41 treatment
with 1 ng/ml exogenous TNF- together with increasing concentrations of -TNF- neutralizing antibody. n 3 independent experiments. ns, not significant; Veh, vehicle. *P
0.05,
****P
0.0001.
LPS MODULATION OF UTERINE MYOCYTE CONTRACTION 7
An alternative mechanism by which procontractile
agonists may act is through the phenomenon of calcium sensitization, whereby the response of the contractile machinery to a given calcium stimulus is enhanced. A major regulator of calcium sensitization is
the small GTPase RhoA, which acts via its effector
ROCK to inactivate myosin light-chain phosphatase
(MLCP), potentiating the effects of elevated intracellular calcium (25). The ROCK inhibitor Y-27632 at 10 M
strongly inhibited basal contraction at 72 h and prevented LPS-mediated enhancement of contraction
(Fig. 7C). Since Y-27632 had a small, but significant,
effect on viability (P
0.01; Fig. 7D), an alternative
ROCK inhibitor was employed to confirm ROCK involvement. GSK-269962 at 50 nM gave very similar
results to Y-27632, dramatically inhibiting basal contraction and abolishing the procontractile effect of LPS
(Fig. 7E), without significantly altering viable cell number
(Fig. 7F). As with AL-8810 at earlier time points, an
interaction effect was observed with LPS treatment, significantly reducing contraction compared to vehicle in the
presence of GSK-269962 (P
0.01), in agreement with a
trend observed for Y-27632 and suggesting that factors
promoting relaxation/quiescence, normally masked by
ROCK activity, are also triggered by LPS in these cells.
Interestingly, the Rho inhibitor CCG-1423, thought to act
via suppression of Rho/megakaryoblastic leukemia 1
(MKL1)-mediated gene transcription (26), was at 1 M
also able to inhibit LPS-induced contraction, but unlike
the ROCK inhibitors, did not significantly alter basal
gel contraction (Fig. 7G). To address whether disruption of RhoA/ROCK pathway signaling was associated
with a general reduction in proinflammatory signaling,
we looked at the effect of GSK-269962 and CCG-1423
treatment on basal and LPS-induced IL-6 and IL-8
secretion by PHM1-41 cells. No effect of either was
observed on IL-6 secretion (Fig. 8A), whereas, interestingly, GSK-269962 (but not CCG-1423) appeared to
enhance LPS-induced secretion of IL-8 (P
0.01, Fig.
8B). Since ROCK-induced MLCP inactivation would be
expected to result in an increase in levels of steady-state
MYL20 phosphorylation, we examined this using in-cell
Western blot analysis to provide an alternative read-out
of LPS effects on contraction. Unadjusted steady-state
Figure 6. A, B) Effect of indomethacin (50 M) on LPS-induced PHM1-41 collagen gel contraction (A) and viability (B) at 72
h. Indomethacin was added concurrently with LPS to generate
CM, as described in Materials and
Methods. n 6 replicates/treatment, representative figure from 1
of 3 independent experiments.
C–E) LPS-induced PHM1-41 collagen gel contraction in the presence of 10 M AL8810 over a 72-h
time course. CM was supplemented with AL8810 on addition to gels. Gel area was recorded at 24 (C). 48 (D), and 72 h (E). F)
Data from panels B–D expressed as a normalized contraction rate. G) Effect of AL8810 on PHM1-41 gel viability at 72 h. n
6 replicates/treatment, representative figure from 1 of 3 independent experiments for each inhibitor. ns, not significant;
Veh, vehicle. *P
0.05 for LPS vs. Veh.
8 Vol. 28 January 2014 The FASEB Journal www.fasebj.org HUTCHINSON ET AL.
levels of phosphorylated MYL20 were significantly increased with LPS treatment at 48 h (P
0.0001; Fig. 8C),
despite an overall reduction in the level of total MYL20
available (P
0.01; Fig. 8D). Correction of phosphorylated against total MYL20 indicates that the proportion
of phosphorylated MYL20 at steady state was increased
by 1.38 0.072-fold with LPS treatment.
DISCUSSION
Organ bath studies using myometrial strips are a standard method of studying uterine contractility, augmented by biochemical studies employing cultured
primary or immortalized uterine myocytes (10, 12, 13,
27). The latter are a particularly valuable tool for
examining myocyte responses in isolation from other
cell types found in myometrium in labor (e.g., macrophages and neutrophils), as well as for longer-term
studies of contractile agonists where ex vivo tissue
viability, viscoelasticity, and sterility may preclude organ
bath studies for 24 h in humans. Uterine myocytes
remain viable and retain a contractile phenotype in
vitro when grown within collagen lattices (10), and in
recent years such collagen gel contraction assays have
been employed to study pharmacological manipulation
of contractile signaling pathways (28). The use of
collagen gel contraction assays also allowed us to circumvent some practical difficulties associated with strategies required to prevent LPS aggregation in aqueous
solution, where the addition of serum (required for
LPS to signal through TLR) results in rapid frothing in
Figure 7. LPS-induced PHM1-41 collagen gel contraction (A, C, E, G) and cell
viability (B, D, F, H) following 72 h
treatment with CM supplemented on
addition to gels with nifedipine (20
g/ml; A, B), Y-27632 (10 M; C, D),
GSK-269962 (50 nM; E, F) or CCG-1423
(1 M; G, H). n 6 replicates/treatment, displaying representative figure
from 1 of 3 each of (nifedipine,
Y-27632) or 1 of 2 (GSK-269962, CCG-
1423) independent experiments for
each ligand. ns, not significant. *P
0.05, **P
0.01, ***P
0.001, ****P
0.0001 vs. vehicle (Veh).
LPS MODULATION OF UTERINE MYOCYTE CONTRACTION 9
an organ bath with standard gas delivery. We reasoned
that the collagen gel contraction assay would provide a
useful alternative format for examining potential modulation of contractility by LPS in more detail.
Here, we demonstrate that myometrial myocytes respond directly to TLR agonists by increasing proinflammatory cytokine production, and, in particular, respond to LPS through TLR4 by contracting. While LPS
does not exert acute (
6 h) effects either on myometrial strip contractility or myocyte MYL20 phosphorylation, it elicits a delayed ( 24 h) contractile response
from primary and immortalized uterine myocytes
grown in collagen gels. LPS induces myocyte production of the procontractile agonist PGF2. PTGS2 and
FP receptor inhibitor studies reveal PGF2-mediated
contractile effects are countered in vitro by a PTGS2-
dependent prorelaxatory factor, with PGE2 and PGI2
being obvious candidates. However, these experiments
also reveal a separate, prostaglandin-independent effect, responsible for the observed increase in contraction at later time points. This effect was abolished by
two separate ROCK inhibitors, and steady-state MYL20
phosphorylation was increased in LPS-treated cells at
72 h, implicating ROCK-dependent calcium sensitization as part of the mechanism of LPS-induced contraction. Notably, basal contraction was also greatly reduced in the presence of these inhibitors, suggesting
that ROCK pathways are a principal determinant of
uterine myocyte tonicity in the context of these in vitro
assays. Inhibition of Rho/ROCK signaling did not
reduce cytokine production by PHM1-41 cells, suggesting that the reduction in LPS-mediated contraction was
not due to dampening of general inflammatory activation. It has recently been demonstrated that both
spontaneous and oxytocin-induced phasic contractions
in human myometrium are associated with rapid
ROCK-dependent cycling of activated MLCP (29),
while ROCK inhibitors have been shown to prolong
gestation in mouse models of PTL (30, 31), highlighting the importance of Rho/ROCK signaling in vivo. In
addition, the ability of the Rho/MKL1 transcriptional
inhibitor CCG-1423 to prevent LPS-mediated contraction suggests that multiple Rho signaling pathways may
contribute to the response. Interestingly, Rho/ROCK
activity is down-regulated by LPS in vascular smooth
muscle cells (32), suggesting that LPS coupling to Rho
pathways is cell-type specific. A summary schematic
showing LPS signaling to contractile pathways (as evidenced by the data in this study) is shown in Fig. 9.
In addition to contraction, PHM1-41 cells produced a
range of inflammatory mediators in response to LPS.
These included the chemokines MIP1, MCP1, IL-8,
CCL5, and CXCL5, as well as G-CSF and GM-CSF, the
labor-associated inflammatory cytokines IL-6 and TNF-,
and the growth factors FGF basic and VEGF. This profile
is supportive of a role for myometrial myocytes in responding to bacterial infection by activating granulocyte
and monocyte recruitment, modulation, and tissue remodeling. The labor-associated cytokine IL-1, which is
elevated in amniotic fluid from patients with infectionassociated PTL (33) and can itself induce contractions
in vivo (34), was notable by its absence. This, together
with the observation that exogenously-added TNF-
stimulated contraction, reinforces the relevance of additional cell types and tissues, such as leukocytes or the
fetal membranes, to the preterm inflammatory environment.
This study demonstrates for the first time that locally
elevated concentrations of bacterial endotoxin may
modulate myometrial contractility independently of
leukocyte recruitment or cytokine production by other
tissues through a direct interaction with myocytes,
providing insight into possible in vivo mechanisms of
PTL. In addition, we show that, in the presence of
TLR4 agonists, tocolysis with nifedipine and/or prostaglandin synthase inhibitors alone may not be successful,
and that targeting the myocyte contractile machinery
may be a more effective therapeutic option in preventFigure 8. A, B) IL-6 (A) and IL-8 (B) production by PHM1-41 cells in response
to LPS, in the absence or presence of GSK-269962 (50 nM) and CCG-1423 (1
M). Supernatants from Fig. 7 were harvested after 24 h for ELISA. C, D)
Phosphorylated MYL20 (C) and total MYL20 (D) levels in PHM1-41 cells
following LPS treatment. Cells were exposed to PBS or LPS for 48 h before
rapid fixation and in-cell Western blot analysis. n 16 replicates/treatment,
representative figure from 1 of 3 independent experiments. Data are expressed
as relative mean fluorescence intensity without further normalization. a
0.01 vs. LPS.
10 Vol. 28 January 2014 The FASEB Journal www.fasebj.org HUTCHINSON ET AL.
ing delivery, ideally in combination with other antiinflammatory agents to prevent infection/inflammation-induced perinatal brain injury.
The authors thank Professor B. Sanborn (Colorado State
University, Fort Collins, CO, USA) for her gift of PHM1-41
cells, Dr A. Henke (University Clinic Cologne, Cologne,
Germany) for his gift of rat tail collagen, the Edinburgh
Reproductive Tissues BioBank for providing myometrial samples, and Professor Philippa Saunders for supporting this
work within the UK Medical Research Council (MRC)–
Centre for Reproductive Health (CRH). This work was
funded by an MRC career development fellowship to J.L.H.
and an MRC Ph.D. studentship to S.P.R., and by grants from
Action Medical Research (SP4275) and Tommy’s.
REFERENCES
1. Goldenberg, R. L., Hauth, J. C., and Andrews, W. W. (2000)
Intrauterine infection and preterm delivery. N. Engl. J. Med. 342,
1500–1507
2. Elovitz, M. A., and Mrinalini, C. (2004) Animal models of
preterm birth. Trends Endocrinol. Metab. 15, 479–487
3. Youssef, R. E., Ledingham, M. A., Bollapragada, S. S.,
O’Gorman, N., Jordan, F., Young, A., and Norman, J. E. (2009)
The role of Toll-like receptors (TLR-2 and -4) and triggering
receptor expressed on myeloid cells 1 (TREM-1) in human term
and preterm labor. Reprod. Sci. 16, 843–856
4. Tribe, R., Moriarty, P., Dalrymple, A., Hassoni, A., and Poston,
L. (2003) Interleukin-1 induces calcium transients and enhances basal and store-operated calcium entry in human myometrial smooth muscle. Biol. Reprod. 68, 1842–1849
5. Herschman, H., Wadleigh, D., and Reddy, S., (2001) Regulation
of COX-2 expression in fibroblasts, osteoblasts, mast cells,
and macrophages. In Advances in Prostaglandin and Leukotriene
Research (Samuelsson, B., Paoletti, R., Folco, G., Granström, E.,
and Nicosia, S. eds.), pp. 41–47, Springer, Houten, The Netherlands,
6. Lorenz, E., Hallman, M., Marttila, R., Haataja, R., and Schwartz,
D. A. (2002) Association between the Asp299Gly polymorphisms
in the Toll-like receptor 4 and premature births in the Finnish population. Pediatr. Res. 52, 373–376
7. Rinaldi, S. F., Hutchinson, J. L., Rossi, A. G., and Norman, J. E.
(2011) Anti-inflammatory mediators as physiological and pharmacological regulators of parturition. Expert Rev. Clin. Immunol.
7, 675–696
8. Thomson, A., Telfer, J., Young, A., Campbell, S., Stewart, C.,
Cameron, I., Greer, I., and Norman, J. (1999) Leukocytes
infiltrate the myometrium during human parturtion: further
evidence that labour is an inflammatory process. Hum. Reprod.
14, 229–236
9. Bell, E., Ivarsson, B., and Merrill, C. (1979) Production of a
tissue-like structure by contraction of collagen lattices by human
fibroblasts of different proliferative potential in vitro. Proc. Nat.
Acad. Sci. 76, 1274–1278
10. Dallot, E., Pouchelet, M., Gouhier, N., Cabrol, D., Ferré, F., and
Breuiller-Fouché, M. (2003) Contraction of cultured human
uterine smooth muscle cells after stimulation with endothelin-1.
Biol. Reprod. 68, 937–942
11. Aguilar, H. N., Zielnik, B., Tracey, C. N., and Mitchell, B. F.
(2010) Quantification of rapid myosin regulatory light chain
phosphorylation using high-throughput in-cell Western assays: comparison to Western immunoblots. PLoS ONE 5,
e9965
12. Anderson, L., Martin, W., Higgins, C., Nelson, S. M., and
Norman, J. E. (2009) The effect of progesterone on myometrial
contractility, potassium channels, and tocolytic efficacy. Reprod.
Sci. 16, 1052–1061
13. Monga, M., Ku, C. Y., Dodge, K., and Sanborn, B. M. (1996)
Oxytocin-stimulated responses in a pregnant human immortalized myometrial cell line. Biol. Reprod. 55, 427–432
14. Slater, D. M., Dennes, W. J. B., Campa, J. S., Poston, L., and
Bennett, P. R. (1999) Expression of cyclo-oxygenase types-1 and
-2 in human myometrium throughout pregnancy. Mol. Hum.
Reprod. 5, 880–884
15. Tattersall, M., Engineer, N., Khanjani, S., Sooranna, S. R.,
Roberts, V. H., Grigsby, P. L., Liang, Z., Myatt, L., and Johnson,
M. R. (2008) Pro-labour myometrial gene expression: are preterm labour and term labour the same? Reproduction 135,
569–579
16. Alexander, H. A., Sooranna, S. R., Myatt, L., and Johnson, M. R.
(2012) Myometrial tumor necrosis factor- receptors increase
with gestation and labor and modulate gene expression through
mitogen-activated kinase and nuclear factor-B. Reprod. Sci. 19,
43–54
17. Yarilina, A., Park-Min, K.-H., Antoniv, T., Hu, X., and Ivashkiv,
L. B. (2008) TNF activates an IRF1-dependent autocrine loop
leading to sustained expression of chemokines and STAT1-
Figure 9. Summary schematic outlining LPS
contractile signaling in uterine myocytes.
LPS MODULATION OF UTERINE MYOCYTE CONTRACTION 11
dependent type I interferon-response genes. Nat. Immunol. 9,
378–387
18. Kawai, T., and Akira, S. (2010) The role of pattern-recognition
receptors in innate immunity: update on Toll-like receptors.
Nat. Immunol. 11, 373–384
19. Li, L.-P., Fang, Y.-C., Dong, G.-F., Lin, Y., and Saito, S. (2012)
Depletion of invariant NKT cells reduces inflammation-induced
preterm delivery in mice. J. Immunol. 188, 4681–4689
20. Goldenberg, R. L., Culhane, J. F., Iams, J. D., and Romero, R.
(2008) Epidemiology and causes of preterm birth. Lancet 371,
75–84
21. Hirschfeld, M., Ma, Y., Weis, J. H., Vogel, S. N., and Weis, J. J.
(2000) Cutting edge: repurification of lipopolysaccharide eliminates signaling through both human and murine Toll-like
receptor 2. J. Immunol. 165, 618–622
22. Huber, M., Kalis, C., Keck, S., Jiang, Z., Georgel, P., Du, X., Shamel,
L., Sovath, S., Mudd, S., Beutler, B., Galanos, C., and Freudenberg,
M. A. (2006) R-form LPS, the master key to the activation ofTLR4/
MD-2-positive cells. Eur. J. Immunol. 36, 701–711
23. Holmgren, C., Esplin, M. S., Hamblin, S., Molenda, M., Simonsen, S., and Silver, R. (2008) Evaluation of the use of anti-TNF-
in an LPS-induced murine model. J. Reprod. Immunol. 78,
134–139
24. Shmigol, A. V., Eisner, D. A., and Wray, S. (1998) Properties of
voltage-activated [Ca2
]i transients in single smooth muscle
cells isolated from pregnant rat uterus. J. Physiol. 511, 803–811
25. Uehata, M., Ishizaki, T., Satoh, H., Ono, T., Kawahara, T.,
Morishita, T., Tamakawa, H., Yamagami, K., Inui, J., Maekawa,
M., and Narumiya, S. (1997) Calcium sensitization of smooth
muscle mediated by a Rho-associated protein kinase in hypertension. Nature 389, 990–994
26. Evelyn, C. R., Wade, S. M., Wang, Q., Wu, M., Iñiguez-Lluhí, J.A.,
Merajver, S. D., and Neubig, R. R. (2007) CCG-1423: a smallmolecule inhibitor of RhoA transcriptional signaling. Mol. Cancer Ther. 6, 2249–2260
27. Norman, J. E., Ward, L. M., Martin, W., Cameron, A. D.,
McGrath, J. C., Greer, I. A., and Cameron, I. T. (1997) Effects of
cGMP and the nitric oxide donors glyceryl trinitrate and sodium
nitroprusside on contractions in vitro of isolated myometrial
tissue from pregnant women. J. Reprod. Fertil. 110, 249–254
28. Eude-Le Parco, I., Dallot, E., and Breuiller-Fouche, M. (2007)
Protein kinase C and human uterine contractility. BMC Pregnancy Childbirth 7(Suppl 1), S11
29. Hudson, C. A., Heesom, K. J., and López Bernal, A. (2012)
Phasic contractions of isolated human myometrium are associated with Rho-kinase (ROCK)-dependent phosphorylation of
myosin phosphatase-targeting subunit (MYPT1). Mol. Hum.
Reprod. 18, 265–279
30. Tahara, M., Kawagishi, R., Sawada, K., Morishige, K., Sakata, M.,
Tasaka, K., and Murata, Y. (2005) Tocolytic effect of a Rhokinase inhibitor in a mouse model of lipopolysaccharide-induced preterm delivery. Am. J. Obstet. Gynecol. 192, 903–908
31. Goupil, E., Tassy, D., Bourguet, C., Quiniou, C., Wisehart, V.,
Pétrin, D., Le Gouill, C., Devost, D., Zingg, H. H., Bouvier, M.,
Saragovi, H. U., Chemtob, S., Lubell, W. D., Claing, A., Hébert,
T. E., and Laporte, S. A. (2010) A novel biased allosteric
compound inhibitor of parturition selectively impedes the prostaglandin F2-mediated Rho/ROCK signaling pathway. J. Biol.
Chem. 285, 25624–25636
32. Wei, C.-Y., Huang, K.-C., Chou, Y.-H., Hsieh, P.-F., Lin, K.-H.,
and Lin, W.-W. (2006) The role of Rho-associated kinase in
differential regulation by statins of interleukin-1- and lipopolysaccharide-mediated nuclear factor B activation and inducible nitric oxide synthase gene expression in vascular smooth
muscle cells. Mol. Pharmacol. 69, 960–967
33. Romero, R., Mazor, M., Brandt, F., Sepulveda, W., Avila, C.,
Cotton, D. B., and Dinarello, C. A. (1992) Interleukin-1 and
interleukin-1 in preterm and term human parturition. Am. J.
Reprod. Immunol. 27, 117–123
34. Sadowsky, D. W., Adams, K. M., Gravett, M. G., Witkin, S. S., and GSK269962A
Novy, M. J. (2006) Preterm labor is induced by intraamniotic
infusions of interleukin-1 and tumor necrosis factor- but not
by interleukin-6 or interleukin-8 in a nonhuman primate model.
Am. J. Obstet. Gynecol. 195, 1578–1589
Received for publication June 21, 2013.
Accepted for publication September 16, 2013.
12 Vol. 28 January 2014 The FASEB Journal www.fasebj.org HUTCHINSON ET AL.