ENHANCED VASCULAR SMOOTH MUSCLE CALCIUM SENSITIVITY AND LOSS OF ENDOTHELIAL VASODILATOR INFLUENCE CONTRIBUTE TO MYOGENIC TONE DEVELOPMENT IN RAT RADIAL UTERINE ARTERIES DURING GESTATION

Abstract

Uterine artery myogenic tone (MT) becomes quite significant during pregnancy in hemochorial placentates such as rats and humans. The physiological reason for its appearance is not clear, and we reasoned that it may be a late pregnancy event in preparation for controlling hemorrhage during parturition. We also hypothesized that gestational increases in RhoA-induced VSM calcium sensitivity are contributory and occur under the tonic influence of nitric oxide. Second order pre-placental radial arteries from early- (EP, day 12, n=5), mid- (MP, day 16, n=5) and late-pregnant (LP, day 20, n=20) rats were used in combination with arteriography, VSM calcium measurements, pharmacological RHO/ROCK and NOS inhibition, and western blotting. L-NAME-treated (LP+LN) animals (gestational days 10-20; n=5) were used to determine the effects of NOS inhibition on MT and RhoA expression. MT was evident throughout pregnancy, but its expression in pressurized vessels was masked by endothelial NO-induced vasodilation during early gestation. RhoA protein expression was upregulated in late pregnancy and attenuated by in vivo NOS inhibition (as was MT). In vitro RHO/ROCK inhibition decreased MT in a concentration-dependent manner without reducing VSM calcium. In summary, pressure-dependent uterine artery tone increases with gestational age due to a combination of RhoA-mediated increases in VSM calcium sensitivity and a loss of endothelial NO influence.

INTRODUCTION

Pressure-dependent arterial tone is essential in blood flow regulation [1, 2] and for protecting smaller downstream vessels from damage due to increased blood pressure. The ambient level of tone reflects an integration of intrinsic vascular smooth muscle (VSM) contractility to pressure or stretch (myogenic tone, MT) and of other, superimposed vasoactive influences, e.g. endothelial vasodilator (e.g. nitric oxide, NO) release.

MT is most often present in smaller resistance arteries and has been documented in vessels from regional circulations other than the uterus, particularly from the brain and kidney [3, 4]. Earlier studies in several regional circulations, including the uterine, have shown that MT results from a synergism between ionic and enzymatic mechanisms that govern VSM calcium entry and calcium sensitivity, respectively [1, 4-9].

The uterine vasculature is unique in that little or no pressure-dependent tone is present in the resistance (radial) arteries outside of pregnancy, but its extent increases considerably during gestation. This adaptation appears to be induced by local uteroplacental influences [10], and has been documented in both rats and humans [10, 11, 21, 26, 27] although, conversely, MT is reduced in the uterine vasculature of the sheep during pregnancy [1, 11-13]. This difference may be related to the type of placentation, as it is important to reduce inflow pressure in hemochorial placentas (human, rodent) to avoid villous damage, while the fetal and maternal vessels are intertwined in epitheliochorial placentas (sheep and other ungulates) in which villi are absent, and placental-inflow pressures are therefore considerably greater [14]. Thus, in view of its hemochorial placentation, the rat is a valuable animal model for understanding hemodynamic changes in human pregnancy.

We hypothesized that: (1) if the appearance of MT would be a late pregnancy (LP) event, and (2) its enhancement during pregnancy arises from a combination of depolarization-induced calcium entry and RhoA-mediated increases in calcium sensitivity in VSM [5, 6]. Based on earlier findings that showed that nitric oxide exerts an important tonic influence on RhoA expression in VSM [15-18] we also hypothesized that NOS inhibition during pregnancy would reduce RhoA protein expression, and that this may account for the as-yet unexplained finding that MT is nearly absent in radial arteries from pregnant rats that underwent NOS inhibition in vivo [19].

METHODS

Animals, Experimental Treatments and Blood Pressure Measurements:

Twelve- to fourteen-week-old timed pregnant Sprague-Dawley (n=30) rats were purchased from Charles River Laboratories (Saint-Constant, QC, Canada) and housed singly in cages at the small animal care facility at the University of Vermont, which is accredited by the American Association for Accreditation of Laboratory Animal Care. All animals were acclimatized for at least 4 days prior to experimentation.

The experimental design reflects the individual animal as an appropriate experimental unit for investigation. All animals were provided feed and water ad libitum under a 12D:12L photoperiod. Early (EP, day 12/22 of gestation, n=5), mid- (MP, day 16/22 of gestation, n=5) and late pregnant (LP, day 20/22 of gestation, n=20) rats were used in this study. Of the LPs, 5 were used for NOS inhibition, 5 for ROCK inhibition, and 5 for Fura-2 calcium measurements.

A fourth subgroup of the LP animals (n=5) was treated with 0.5 g/L L-NAME in drinking water from gestational days 10-20 (as in earlier studies), and designated as LP+LN [19]. Daily water intake per animal was measured on every other day for the latter, and blood pressures were measured noninvasively by a tail-cuff method (CODA System, Kent Scientific, Torrington, CT.) on the morning of each experimental day prior to euthanizing the animal for tissue collection to assure sufficient delivery of the drug, and to provide a readout of its hypertensive effects. Rats were not habituated to blood pressure measurements, as they were used for confirmatory rather than comparative purposes. Twenty readings were taken from each animal, and averaged.

Vessels used for Western Blotting were collected in addition to those used for pressure myography, hence, no additional animals were required.

Each rat was anesthetized with 3% isoflurane, weighed and euthanized by decapitation in a small animal guillotine. The uterus and its vasculature were removed in its entirety and pinned in a silicone-coated Petri dish filled with freshly prepared, room temperature HEPES-physiologic salt solution (PSS). All experimental protocols were approved by the Institutional Animal Care and Use Committee (IACUC) in accordance with NIH and ARRIVE guidelines.

Isolated Vessel Preparation:

Proximal segments of second-order pre-placental radial arteries [20] were dissected free from perivascular adipose tissue. The ends of each vessel were cannulated in an arteriograph (Living Systems Instrumentation, St. Albans, VT) superfused with HEPES-PSS at 37°C, pressurized to 60 mmHg and equilibrated for 45 min under no-flow conditions. Transmural pressure and lumen diameter were recorded during each experiment using an IonOptix data acquisition program (Westwood, MA).

Experimental procedure:

Following equilibration at 60 mmHg, intraluminal pressure in each cannulated vessel was increased to 80, and then 100 mmHg. Each pressure step was maintained for 15-20 min to allow the vessel to reach a stable diameter. Pressure was then decreased to 60 mmHg and L-NAME (100μM) plus N(ω)-nitro-L-arginine (L-NNA, 100μM) were added to the superfusate for 20 min to induce nitric oxide synthase (NOS) inhibition prior to repeating the pressure stair.

Fully relaxed (passive) lumen diameters were obtained at the end of each experiment by superfusing each vessel for 20 min with a relaxing solution containing 10 μM diltiazem and 100 μM papaverine; n values represent the number of animals, with one vessel from each used for reactivity experiments, and several others collected for Western blot analyses. In the interest of clarity, 60 mmHg, which approximates the physiologic pressure most likely experienced by uterine second order radial arteries [14] was used in subsequent ROCK inhibition and calcium measurement experiments.

Effects of p160 Rho-associated protein kinase (ROCK) inhibition with Y27632 on MT:

The effects of Y27632 were tested (3 and 10 μM applied cumulatively; 20 min at each concentration) post-equilibration at 60 mmHg (with tone) in vessels from 5 animals.

Fura-2 Loading and Smooth Muscle Cell (SMC) Cytosolic Calcium [Ca²⁺]_i Measurement in Pressurized Arteries:

Ratiometric measurements of fura-2 fluorescence from smooth muscle cells in the vascular wall (media) were performed using a photomultiplier system (IonOptix Inc. Milton, MA). Arteries were cannulated and pressurized to 60 mmHg. Background fluorescence was measured at 37°C from each artery before loading with 5 μM fura2-AM. 10 μl of fura 2-AM stock solution in dehydrated DMSO was premixed with 10 μl of a 20% solution of pluronic acid and added to 2 ml of HEPES-PSS. This solution was added to the superfusate in an arteriograph containing a radial artery pressurized to 10 mmHg for 60 min at room temperature and no circulation in the chamber to allow Fura2-Am to penetrate into the media (vascular smooth muscle). After loading, arterial segments were superfused for 15 min at 10 mmHg and 37°C to allow intracellular de-esterification of fura-2AM. Background-corrected ratios of 510 nm emission were obtained at a sampling rate of 5 Hz from arteries alternately excited at 340 and 380 nm.

Following warming to 37°C, intraluminal pressure was increased from 10 to 60 mmHg. The effects of Y27632 (3, 10 μM) on MT were tested during a 10 min exposure at each concentration. Transmural pressure, lumen diameter and 340/380 fura-2 ratios were recorded throughout the experimental protocol. At the end of each experiment, each vessel was superfused with a relaxing solution of HEPES containing papaverine (100 μM) + diltiazem (10 μM) to achieve maximal relaxation as required for calculation of tone.

Western Blot Analysis:

Following uterine excision, pre-placental radial arteries from EP, LP and LP+LN animals were collected in addition to those used for pressure myography and immediately frozen in liquid nitrogen, then kept at −80°C until the day of analysis.

Arteries were lysed with Pierce RIPA buffer (ThermoFisher Scientific, Rockford, IL) supplemented with Halt protease inhibitor cocktail (ThermoFisher Scientific) and homogenized in Lysis Matrix D tubes (MP Biomedicals, Solon, OH) using two 30-s pulses on a FastPrep-24™ homogenizer (MP Biomedicals). Total protein concentrations were determined by BSA assay (ThermoFisher Scientific). Protein samples (3 μg of soluble protein each) were separated using 4-15% Criterion™ TGX™ precast gels (Bio-Rad Laboratories, Hercules, CA) and transferred to a PVDF membrane. After blocking with 5% nonfat milk, rabbit monoclonal eNOS (1:1000, Cell Signaling Technology, Denvers, MA), rabbit polyclonal β-tubulin (1:5000, Cell Signaling Technology), and mouse monoclonal RhoA (Cytoskeleton Inc., Denver, CO) primary antibodies and horseradish peroxidase-conjugated secondary antibodies (Abcam, Cambridge, UK) were used to detect specific proteins on the blot. Protein bands were detected by SuperSignal West Pico Plus chemiluminescent substrate (ThermoFisher Scientific) and analyzed by densitometry using ImageJ software.

Drugs and Solutions:

HEPES-PSS contained: 141.9 mM NaCl, 4.7 mM KCl, 1.7 mM MgSO₄, 2.8 mM CaCl₂, 0.4 mM EDTA, 10.0 mM HEPES, 1.2 KH₂PO₄, and 5.0 dextrose, pH=7.4 at 37°C. The Fura-2 calibration solution contained: 140 mM KCl, 20 mM NaCl, 5 mM HEPES, 5 mM EGTA, 1 mM MgCl₂ 5 μM nigericin and 10 μM ionomycin, pH = 7.4 at 37°C. All solutions were prepared on the day of an experiment.

Chemicals were purchased from Fisher Scientific (Fair Lawn, NJ) unless stated otherwise. Ionomycin (Alfa Aesar, Ward Hill, MA) and nigericin (Calbiochem, La Jolla, CA) were prepared as 10 mM stock solutions in DMSO and kept at −20°C until use. Fura-2 (Invitrogen, Carlsbad, CA) was dissolved in DMSO as a 1 mM stock solution, refrigerated in small aliquots and used within 1 week of preparation. L-NAME, papaverine and diltiazem were purchased from Sigma-Aldrich (St. Louis, MO) and L-NNA from Alfa Aesar (Haverhill, MA); Y27632 (hydrochloride) was purchased from Cayman Chemical (Ann Arbor, MI), dissolved in DMSO as a 10 μM stock solution and frozen at −20°C until use.

Calculations and Statistical Analysis:

MT (%) was calculated using the following equation: [(D_P−D_A)/D_P] x 100, where D_P is passive diameter obtained in response to papaverine and diltiazem, and D_A is active diameter of the pressurized vessels. GraphPad Prism 7.01 was used for graphical representation, calculation, and statistical analysis of data.

SMC [Ca²⁺]_i was calculated using the equation: [Ca²⁺]_i = K_dβ (R – R_min)/(R_max - R), where R is the experimentally measured ratio (340/380 nm) of fluorescence intensity; R_min is a ratio in the absence of [Ca²⁺]_i ; R_max is a ratio at Ca²⁺-saturated fura-2 conditions, and β is a ratio of the fluorescence intensities at 380 nm. Excitation wavelengths at R_min, R_max and β were determined by an in situ calibration procedure from arteries treated with nigericin (5 μmol/L) and ionomycin (10 μM) [10]. Calibration was performed on a group of vessels loaded extraluminally with fura-2 (n=4). These values were then pooled and used to convert the ratiometric values into actual [Ca²⁺]_i . K_d (the dissociation constant for fura-2) was 282 nM [20].Transmural pressure, lumen diameter and ratio values were simultaneously recorded using an IonOptix data acquisition program and imported into SigmaPlot programs for graphical representation, calculation, and statistical analysis.

Data are expressed as mean ± SEM. Unpaired Student’s t-test and one-way RM ANOVA were used to determine the significance of differences between groups, as were non-repeated measures ANOVA (one-way and two-way, based on number of independent variables; see figure legends). Statistical differences were considered significant at P < 0.05.

RESULTS

Time course and extent of MT development during rat pregnancy; effects of NOS inhibition:

Pressure-dependent tone at 60 mmHg was minimal in arteries from EP (1.5 ± 0.52%) and MP (3.1 ± 0.57%) animals, but considerable (35 ± 5.4%) in LP vessels (n=5 /group, p<0.05; Figure 1A). As shown in Figure 1B, tone was somewhat increased at higher pressures (80, 100 mmHg) and NOS inhibition (L-NAME + L-NNA added to the superfusate) significantly enhanced pressure-dependent tone in EP but not MP or LP vessels.

Figure 1:

[A] Pressure-dependent tone (%) in isolated, pressurized (60 mmHg) pre-placental second-order radial arteries from early- (EP, day 12/22 of gestation), mid- (MP, day 16/22 of gestation) and late-pregnant (LP, day 20/22 of gestation) animals. The level of tone was significant in arteries from LP animals and essentially absent in those from the EP and MP groups (n=5 /group). [B] Summary graph showing the level of pressure-dependent tone as a function of gestational age and transmural pressure in the absence (open bars) vs. presence (shaded bars) of in vitro NOS inhibition. A significant increase in MT following NOS inhibition was only observed in vessels from EP animals at 80 and 100 mmHg, although an effect was clearly seen at 60 mmHg as well. Data are expressed as means ± SEM. Asterisk denotes statistical significance relative to each of the other groups, p<0.05 (n=5/group, one way ANOVA) in Figure 1A; in Figure 1B, comparisons between control and NOS-inhibited vessels were made within each gestational age and at each transmural pressure (two-way repeated measures ANOVA). * = p < 0.05; ** = p < 0.01.

Effects of ROCK inhibition on MT and VSM cytosolic calcium:

The application of 10 μM Y27632, a potent inhibitor of p160ROCK [21], to vessels from LP rats reduced the extent of MT by 85 ± 7.0% (p<0.05), with intermediate vasodilation evident in response to the 3 μM concentration as well (Figure 2).

Figure 2:

Rho kinase inhibition by 10 μM Y27632 significantly reduced MT in pre-placental second-order radial arteries (60 mmHg) from late-pregnant animals. Some vasodilation occurred at the 3 μM concentration as well, but this was not significantly different from controls. Data are expressed as means ± SEM. Asterisk denotes statistical significance relative to the other groups, p<0.05 (n=5 /group, one-way repeated measures ANOVA).

This loss of tone was not associated with any decreases in vascular SMC cytosolic [Ca²⁺]_{i .} Rather, as shown in the tracing (Figure 3A), and in the data summary (Figures 3B and C), [Ca²⁺]_i increased progressively in response to each concentration of Y27632, significantly so at the higher (10 μM) concentration.

Figure 3:

Inhibition of Rho kinase (ROK) induced a concentration-dependent increase in cytosolic Ca²⁺ concentration [Ca²⁺]_i and loss of myogenic tone (dilation) in pre-placental second-order radial arteries. A: Representative experimental tracings showing the effects of Y27632 (3 and 10 μM) on vascular smooth muscle cell (VSMC) cytosolic [Ca²⁺ ]_i and lumen diameter in one vessel from an LP rat. Note rise in calcium during the initial development of myogenic tone following the increase in intraluminal pressure from 10 to 60 mmHg. Once vascular diameter and [Ca²⁺ ]_i stabilized, the addition of 10 μM Y27632 produced modest increases in [Ca²⁺ ]_i and loss of tone in a concentration-dependent manner. The dotted lines indicate basal calcium (middle trace) and the relaxed lumen diameter (lower trace) of the same artery in the presence of 100 μmol/L papaverine and 10 μmol/L diltiazem. As noted earlier [28], vasomotion is characteristic of uterine arteries from pregnant rats. B: Summary graphs showing VSMC cytosolic [Ca²⁺ ]_i (top; calcium increased significantly in 10 μM Y27632 vs. control), and the significant decrease in % MT (bottom) in response to 10 μM Y27632 in pre-placental second-order radial arteries pressurized to 60 mmHg. The effects of 3 μM Y27632 were intermediate in both cases. Data are expressed as means ± SEM. Asterisk denotes statistical significance, p<0.05 (n=5 /group, one-way repeated measures ANOVA).

Effects of in vivo NOS inhibition on mean arterial blood pressures and myogenic tone:

Mean arterial pressures (MAPs) at the end of the 10-day treatment with L-NAME (0.5 g/L) averaged 119 ± 6.6 mmHg (n=5) in LP+LN animals, some 20 mmHg higher than that MAPs normally measured in control animals [19].

In vivo NOS inhibition with L-NAME from gestational days 10 to 20 prevented the development of significant MT in radial arteries from LP animals (n=5 /group, p<0.05; Figure 4); the level of tone in LP-LN vessels was reduced by >80% compared with controls.

Figure 4:

Myogenic tone was significantly (>80%) reduced in pressurized pre-placental second-order radial arteries from NOS-inhibited (L-NAME, 0.5 g/L from days 10-20 of gestation) vs. control LP rats. Data are expressed as means ± SEM. Asterisk indicates statistical significance relative to LP control, p<0.05 (n=5 /group, unpaired Student’s t-test)

RhoA protein expression in vessels from EP and LP control, and from LP NOS-inhibited (LP+LN) rats:

RhoA protein was detectable in vessels from EP animals, and its expression was significantly increased in vessels from LP animals (Figure 5). RhoA protein also increased significantly in radial arteries from rats that underwent NOS inhibition with L-NAME in vivo (from days 10-20 of pregnancy), but to a level that was between that of vessels from EP and LP animals, and significantly different from both (Figure 5).

Figure. 5:

Comparison of RhoA protein expression in pre-placental second-order radial arteries from early (EP, day 12/22 of gestation), late-pregnant (LP, day 20/22 of gestation), and late-pregnant L-NAME-treated (LP + LN) rats. Top: Western blot gel showing RhoA and β-tubulin protein expression in vessels from EP, LP and LP+LN rats; Bottom: densitometric analysis normalized to β-tubulin. Compared to arteries from LP rats, RhoA protein expression was significantly lower in vessels from EP and LP+LN groups. Data are expressed as means ± SE. Asterisk denotes statistical significance, p<0.05 (n=4/group, one-way ANOVA)

DISCUSSION

Effects of gestational age and transmural pressure on pressure-dependent vs. myogenic tone:

The physiological purpose underlying the appearance of uterine artery pressure-induced tone in late pregnancy is not known but may have to do with increasing flow resistance in order to prevent excessive downstream (spiral artery) pressures which could otherwise lead to accelerated placental inflow and fetal villous damage [22] and/or possibly limiting hemorrhage during parturition.

As shown in Figure 1, pressure-dependent tone at 60 mmHg was minimal in vessels from EP and MP animals and significant in those from LP rats. At higher pressures (80 and 100 mmHg, Figure 2), pressure-dependent tone was increased somewhat in arteries from EP and MP animals as well, although the values were significantly lower than those of arteries from LP animals.

Interestingly, the in vitro NOS inhibition data in Figure 1B show that significant myogenic tone is present in early pregnancy as well, but is almost completely inhibited by the vasodilatory influence of endothelial NO, since NOS inhibition increased tone at every pressure, and significantly so at 80 and 100 mmHg. This overriding effect of NO was no longer present by mid-pregnancy. Thus, the capacity of VSM to develop MT does not appear to be a late-pregnancy phenomenon; rather, VSM is capable of myogenic responsiveness throughout pregnancy, but its expression earlier in pregnancy is masked by the overriding effect of endothelial NO. This influence is reduced and eventually lost as pregnancy progresses such that considerable MT is present in vessels from LP rats, and the effects of NOS inhibition become insignificant.

Species differences in myogenic tone (MT) during pregnancy:

With the exception of one report in mice [23], most published studies on rodents, rabbits and humans have documented increases in uterine artery MT during pregnancy [19, 20, 24-28]. This is quite different from the situation in the ewe, in which uterine artery MT during gestation is decreased. One explanation for the species differences may relate to the type of placentation. Ungulates such as sheep have epitheliochorial placentas, in which the maternal vascular network is preserved, and intertwines with the fetal placental vessels so that the exchange of nutrients and wastes occurs from vessel to vessel. Accordingly, intravascular pressures in the small maternal arteries are reportedly close to systemic in epitheliochorial placentates (84 mmHg, [14]), whereas in species with hemochorial placentation (such as humans, rats and guinea pigs), they are on the order of only 9-14 mmHg [14], since the placenta essentially forms a chamber that is perfused with maternal blood in a high volume, low velocity flow pattern. Since blood pressures in the main uterine arteries are close to systemic [14], a considerable pressure drop occurs along the length of the radial arteries, and the presence of MT thus provides an effective mechanism for regulating flow resistance in a bidirectional manner.

Physiological mechanisms responsible for altering uterine artery MT during gestation:

Our experimental results are consistent with previously published studies in the uterine, and other regional circulations (e.g. cerebral) showing that pressure-induced MT arises from a combination of ionic and enzymatic mechanisms that regulate cytosolic calcium and calcium sensitivity, respectively. As shown in the uterine, cerebral and renal circulations, pressure-induced VSM depolarization leads to calcium entry through L-type voltage-dependent calcium channels [29]. The contractile effect of increased cytosolic calcium is further augmented by the activation of calcium-sensitizing enzymes such as protein kinase C (PKC) and/or RhoA which alter the balance between myosin light chain kinase (MLCK) and phosphatase (MLCP) activity in a way that favors actomyosin interaction [4, 30-32].

For example, Zhang and colleagues have demonstrated a number of underlying vascular adaptations that underlie the gestational reduction in MT in uterine arteries from the ewe. These include: (1) Upregulation of large conductance calcium-activated potassium (BK_Ca) channel activity [11], (2) increased calcium sparks and spontaneous outward currents (STOCs) which activate (BK_Ca) channels and elicit VSM hyperpolarization [11], (3) downregulation of actin polymerization [12], and (4) a reduction in calcium sensitivity due to an attenuation of protein kinase C/extracellular signal-regulated kinase (PKC/ERK) signaling [18, 33]. Many of these changes can be mimicked in nonpregnant animals by chronic treatment of nonpregnant tissues with 17beta-estradiol and progesterone, underscoring the role of sex steroids in their genesis.

As already discussed, in rats, rabbits and humans, the consensus is that pressure-induced MT is increased in pregnancy [19, 20, 24-28] and that both calcium handling and calcium sensitivity mechanisms such as decreased function (rather than expression) of voltage-gated potassium (K_v) channels [28] and upregulation of VSM RhoA signaling as a means for increasing VSM calcium sensitivity (this study).

One report in mice [23] described a reduction in MT in the main uterine artery and attributed it to a combination of increased endothelial nitric oxide release and gap junctional communication. In our experience, there is no tone in the main uterine artery of the rat (data not shown) but there are functional differences based on arterial location [10,20]; clearly, additional studies are needed to clarify differences in arterial tone based on vascular architecture and species.

In terms of the molecular mechanisms underlying uterine artery tone development during pregnancy, less is known about rodents than sheep. Using a model of unilateral oviductal ligation (which results in a ‘half-pregnant’ rat since implantation occurs in only one uterine horn), we found that MT was only present in vessels in the implanted horn, suggesting that local uteroplacental influences may be responsible [10]. If, as in sheep, sex steroids are involved (especially progesterone, since estrogen is produced in the ovaries and not the placenta of a pregnant rat, [34, 35]), this raises the question of why the effect is localized to only the implanted horn, since arterial steroidal concentrations are increased systemically and should therefore equally affect vessels in both uterine horns.

One as-yet putative pathway is venoarterial communication, a process in which placental signals secreted into the uterine vein pass across the venous wall into the periarterial mesometrial space to modulate arterial structure and tone. We recently found that redirecting the venous outflow by removing a segment of vein next to an artery significantly reduced the expansive (outward) remodeling in the adjacent arterial segment [36]. This is the first in vivo evidence implicating venoarterial communication in the process of maternal uterine arterial remodeling. It also raises the possibility that the process of arterial expansive remodeling itself may alter VSM function, e.g. MT may be induced through changes in gene expression that occur during the remodeling process.

Mechanisms underlying upregulation of RhoA signaling

This is the first paper to show that RhoA signaling is largely responsible for MT in the rat radial artery of LP rats, since its inhibition by incubation with Y27632, a selective p160ROCK inhibitor, reduced MT by >80%. Although the mechanisms through which RhoA is upregulated during pregnancy in the maternal uterine circulation unknown, several studies by Loirand and co-workers [15-17] deserve mention because they showed that RhoA expression in VSM is tonically regulated by endothelial NO.

Specifically, in vivo treatment with L-NAME significantly reduced RhoA signaling in rat aortas and pulmonary arteries, and this effect was prevented by oral co-administration of sildenafil [15-17]. In rat or human arterial smooth muscle cells, sodium nitroprusside or 8-(2-chlorophenylthio)-cGMP induced a rise in RhoA mRNA and protein expression, and this could be inhibited by (R(p))-8-bromo-bet-phenyl-1,N(2)-ethenoguanosine 3’:5’-phosphorothioate, a PKG inhibitor. In a subsequent study [17], treatment with the phosphodiesterase-5 inhibitor sildenafil restored and augmented RhoA substantiating VSM cGMP-PKG pathway as the responsible pathway. These effects of NO-cGMP signaling were linked to both stimulation of RhoA gene transcription and a post-transcriptional increase in RhoA protein stability.

With respect to our results on the effects of gestation, this is quite plausible since NO signaling is augmented in pregnancy on a systemic level [37], and locally in the uterine circulation, as first shown by the 8-fold increases in calcium-dependent NOS activity (and expression) in uterine arteries from pregnant vs. nonpregnant women [38, 39]. Many subsequent studies by Magness and Bird [reviewed in [40] and [41]] and others, including our own [14, 37] have confirmed the importance of NO signaling in mediating gestational changes in uterine artery reactivity and remodeling.

Here, in vivo NOS inhibition with L-NAME from days 10 to 20 of pregnancy significantly reduced RhoA protein expression within the uterine artery wall and reduced the level of arterial tone by >80%, as in earlier studies [19,20]. The RhoA effect was intermediate, i.e. based on the Western blotting intensity analysis, RhoA expression approximately tripled in vessels from LP control animals, but only doubled in rats treated with L-NAME (0.5 g/L, added to drinking water).

Limitations. considerations

The use of isolated, pressurized vessels eliminates a number of factors extrinsic to the vascular wall that may impact on the extent of arterial tone present in vivo, such as humoral/hormonal or neural influences, and physical forces such as pulsatile intravascular pressure and the shear stress of blood flowing over the endothelial luminal surface.

Likewise, although endothelial NO is probably the major influence on uterine artery VSM, L-NAME inhibition is not specific to NOS3 (eNOS), and it is therefore not possible to attribute the changes in MT and in RhoA expression specifically to an altered influence of endothelial NO. Many other cell types generate NO and these may, in turn, affect the arterial wall via direct paracrine action (e.g. macrophages, nerves or mast cells), or indirectly (e.g. through changes in intravascular pressure). Although we inhibited NO signaling, endothelial denudation was not included in the experimental design, and some endothelial influence from other endothelium-derived vasoactive molecules cannot be excluded.

Although NOS inhibition has been used as an animal model for human preeclampsia, like most animal models, it only approximates human disease. It would be interesting to examine uterine artery MT in another animal model of preeclampsia (such as sFlt-1or STOX1 overexpression) or, better yet, in human intra-myometrial vessels obtained from normal vs. preeclamptic women during cesarean section, as in earlier studies [24, 27].

We also do not know the extent to which L-NAME inhibited the gestational upregulation of NO signaling in the uterine circulation. Measuring this parameter in vivo accurately is difficult since blood or urinary measurements reflect systemic and not local (uterine) changes, and a number of factors that modulate NOS activation (e.g. estrogen, endothelial cytosolic calcium, tetrahydrobiopterin, oxidative state etc.) are both interactive and altered during mammalian gestation.

In summary, the findings of this study delineate the extent and basis of uterine resistance (radial) artery pressure-dependent MT as a function of gestational age and show that increases in Rho-A signaling may contribute to the manifestation of tone by increasing vascular smooth muscle calcium sensitivity. They also indicate that MT is obscured by NO during early gestation, and that this influence decreases with gestational age. The importance of MT in the maternal uterine circulation during pregnancy is not known, but we one possibility is that it may have to do with controlling hemorrhage during parturition, which would explain why its appearance is a late pregnancy phenomenon.