- Email: [email protected]
Endogenous modulation of the blunted adrenergic response in resistance-sized mesenteric arteries from the pregnant rat
Endogenous modulation of the blunted adrenergic response in resistance-sized mesenteric arteries from the pregnant rat Sandra T. Davidge, MS, and Margaret K. McLaughlin, PhD
Cincinnati, Ohio OBJECTIVE: We tested the hypothesis that during pregnancy the endothelium mediates the blunted response to adrenergic vasoconstriction. STUDY DESIGN: Mesenteric resistance arteries from late pregnant (n = 6) and age-matched virgin control (n = 6) Sprague-Dawley rats were studied in a myograph. RESULTS: Arteries from pregnant rats were 35% less sensitive to phenylephrine vasoconstriction than were those from nonpregnant rats (mean effective concentration that produced a 50% response 2.26 vs 1.48 Il-mol/L, pregnant vs nonpregnant, p < 0.01). Meclofenamate had no effect on the vasoconstrictor response in arteries from either group. Inhibition of endothelium-derived relaxing factor with W-nitro-L-arginine methyl ester or endothelial cell removal had a similar twofold increase in phenylephrine sensitivity in arteries from both the pregnant and nonpregnant rats (mean effective concentration that produced a 50% response 2.26 vs 1.11 Il-mol/L for pregnant rats and 1.48 vs 0.72 Il-mol/L for nonpregnant rats, p < 0.01). However, methacholine relaxation response was potentiated in pregnant versus nonpregnant rats (mean effective concentration that produced a 50% response 0.030 vs 0.049 Il-mol/L, p < 0.01). CONCLUSION: Although the potential for endothelium-dependent relaxation is augmented in mesenteric arteries of the pregnant rat, the decreased sensitivity to phenylephrine during pregnancy is not modulated acutely by endothelium-derived relaxing factor or by prostaglandin products of the cyclooxygenase pathway. (AM J OSSTET GYNECOL 1992;167:1691-8.)
Key words: Pregnancy, adrenergic vascular reactivity, resistance arteries, endothelium, Sprague-Dawley rat The maternal hemodynamics of normal pregnancy have been well described. This includes the decrease in peripheral vascular resistance that is associated with a decreased pressor responsiveness to vasoconstrictors. I This reduction in vascular reactivity is clinically important because it is not evident in women with pregnancyinduced hypertension or preeclampsia. 2 In spite of considerable research the mechanisms that could account for this decreased pressor responsiveness during pregnancy are not well understood. Specific alterations within the vascular wall are likely to contribute significantly to this gestational change in vascular reactivity. The endothelium produces and releases substances such as prostacyclin and endothelium-derived From the Perinatal Research Institute, Division of Neonatology, Department of Pediatrics, University of Cincinnati College of Medicine. Supported in part by United States Public Health Service grant No. 40130.
Presented at the Thirty-ninth Annual Meeting of the Society for Gynecologic Investigation, San Antonio, Texas, March 18-21, 1992. Reprint requests: Margaret K. McLaughlin, PhD, Children's Hospital Medical Center, TCHRF - Neonatology Division, Eiland and Bethesda Ave., Cincinnati, OH 45229-2899. 6/6/41903
relaxing factor (EDRF) that will modulate the constriction of vascular smooth muscle. 3 However, it remains controversial if these vasorelaxants are involved in the decreased vasoreactivity observed during pregnancy.<-12 One reason for the conflicting data may involve the use of conduit arteries from different vascular beds. Therefore we have chosen to study resistance-sized mesenteric arteries because this vascular bed is a major determinant of total peripheral vascular resistance. This study was designed to test the hypothesis that the reduction in adrenergic reactivity that occurs in resistance-sized mesenteric arteries during pergnancy13 is in part modulated by endogenous vasorelaxants of the vessel wall. In particular, we were interested in studying the differential influence of EDRF and products of the cyclooxygenase pathway in modulating the vasoconstrictor response to phenylephrine in arteries from pregnant and nonpregnant rats.
Methods General animal model. Ten-week-old female Sprague-Dawley rats were obtained from Harlan (Harlan, Ind.) and bred in our own colony. Each morning vaginal smears were examined microscopically, and the
1691
1692 Davidge and Mclaughlin
presence of sperm was considered day 1 of gestation (term 22 days). The experiments were performed during late (18 to 19 days) gestation and in age-matched virgin controls. The animals were housed in the facilities of the Department of Laboratory Animal Medical Services at the University of Cincinnati, which is accredited by the American Association for the Accreditation of Laboratory Animal Care. Purina 5001 Rat Chow and water were provided ad libitum. Vessel preparation. Rats were decapitated while they were under light anesthesia with methohexital sodium (50 mg/kg body weight). A section of the mesentery 5 to 10 cm distal to the pylorus was rapidly removed and placed in ice-cold N-[2-hydroxyethyl]piperazine-N' -[2ethanesulfonic acid] (HEPES)-buffered physiologic saline solution. One mesenteric artery averaging 250 fLm in diameter was dissected free from surrounding adipose tissue, cut into two 1.8 cm lengths, and threaded onto 16 fLm wires. These wires were fastened to two stainless steel blocks that were mounted in a specially designed myograph system. 14 One block was attached to a Kulite strain gauge force transducer (Kulite Semiconductor Products, Ridgefield, N.J.), and the other was connected to a displacement device. The blocks rested in 10 ml glass-jacketed organ baths with HEPES-physiologic saline solution, kept at 37° C. Each experiment had two baths running simultaneously. Resting length-tension curve. Mter being mounted, the arteries were stretched to about 0.2 nM/mm vessel length (1 mN = 102 mg) and allowed to equilibrate for 1 hour in HEPES-physiologic saline solution buffer. The arteries were then given a conditioning stretch of about 0.6 mN. To generate the resting curve, each artery was stretched in six incremental steps and was held for 20 seconds at each step. The force at each stretch was read at the end of 20 seconds, and the displacement was measured. We have termed this the resting length-tension relationship. The arterial circumference that was used to perform the dose-response curves was obtained by using the Law of LaPlace. With this equation LIOO is calculated from the exponential curve fit of tension versus circumference. LIOO is defined as the circumference the vessel would have at a transmural pressure of 100 mm Hg. We have found from our previous studies that a doseresponse curve obtained at 0.8 LIOO is a point that provides maximum active force generation with minimum passive tension in arteries from both nonpregnant and pregnant female rats. Solutions and drugs. The HEPES-physiologic saline solution used in these experiments contained sodium chloride 142 mmol!L, potassium chloride 4.7 mmol!L, magnesium sulfate 1.17 mmol!L, calcium chloride 1.56 mmol/L, potassium phosphate 1.18 mmol!L, HEPES 10 mmol/L, and glucose 5.5 mmol!L. The HEPES-physi-
December 1992 Am J Obstet Gynecol
ologic saline solution was maintained at a pH of 7.4. Stock solutions of phenylephrine (L-phenylephrine hydrochloride), methacholine (acetyl-l3-methylcholine chloride), sodium nitroprusside (all from Sigma, St. Louis), and meclofenamate (Warner Lambert, Ann Arbor) were prepared in HE PES-physiologic saline solution at a concentration of 10 mmol!L for each experiment. A stock solution of 10 mmol!L of Nw-nitro-Larginine methol ester (LNAME, Sigma) was prepared in water. Appropriate dilutions of all stocks were obtained with HEPES-physiologic saline solution. Experimental design. Vasoreactivity was measured with the aI-adrenergic agonist phenylephrine. To determine the role of endogenous vasorelaxants in modulating the vasoconstrictor response to phenylephrine, meclofenamate was administered to block cyclooxygenase conversion of arachidonic acid and LNAME was used to block the conversion of L-arginine to EDRF. The order of drug administration and the time between drug exposure for these experiments were determined in a preliminary set of experiments designed to test for tachyphylaxis and drug interaction. Cumulative doses of phenylephrine (0.3 to 10 fLmol! L) were administered 30 minutes after the completion of a resting-length tension curve. Mter completion of the dose-response curve, a 30-minute recovery period was allowed, during which the baths were frequently changed with fresh HEPES-physiologic saline solution. By means of two separate baths, one segment of the divided artery from each rat had the phenylephrine dose-response curve repeated in the presence of meclofenamate (1 fLmol!L), whereas the other segment had the phenylephrine dose-response curve repeated in the presence ofLNAME (0.25 mmol!L). Meclofenamate was added 30 minutes before and LNAME 10 minutes before the repeat phenylephrine response curve was generated. At the end of each experiment the arteries were preconstricted with phenylephrine to 50% of their maximal response, and a single dose of methacholine (1 fLmol!L) was administered to prove that a functional endothelium existed for these experiments. A second set of arteries from the same rats were denuded of endothelium before the administration of cumulative doses of phenylephrine (0.3 to 10 fLmol!L). Endothelium removal was done mechanically with a human hair threaded through the lumen of the artery and rubbed back and forth.15 Confirmation of complete endothelium removal was done pharmacologically with a single dose of 1 fLmol!L methacholine at the end of the experiment. Previous experiments with silver staining and scanning electron microscopy had shown that pharmacologic confirmation of endothelial cell removal is appropriate for these arteries (unpublished observation). In a separate protocol with mesenteric arteries from a
Endogenous modulation of vasoconstriction
Volume 167 Number 6
100
• II
C
0
a.
100
•c II
80
0
a.
•
II
0::
E E
0::
60
~
0
....c
•f:! •0..
E E
60
:::I
40
::J
'xg
:::I
80
•
II
::J
1693
'xg
40
~
0
....c
•f:! •
20
0..
20
0
0
10- 8
10-5
Phenylephrine (mil) Fig. 1. Concentration-response curves to phenylephrine for mesenteric arteries from pregnant (n = 6, solid triangles) and nonpregnant (n = 6, solid circles) rats. Tension is expressed as a percentage of maximum phenylephrine response. Data represent mean ± SE.
different group of rats, a cumulative dose-resonse curve to the endothelium-dependent vasorelaxant methacholine (10 nmo1!L to 1 fLmo1!L) and the endotheliumindependent vasorelaxant sodium nitroprusside (1 nmo1!L to 0.1 fLmo1!L) was obtained after preconstricting the arteries with phenylephrine to 50% of the maximum response. Data analysis. The data from the dose-response curves were fitted to the Hill equation,16 from which a straight line was generated by linear least-squares regression analysis. The mean effective concentration that produced a 50% response (EC so ) was determined from this line and expressed as the geometric mean ± SE. Dose ratios were determined as the ratio of the EC so before treatment to that after treatment. Comparisons between groups and treatments were done with a twoway analysis of variance with repeated measures. A value of p < 0.05 was considered significant.
Results The pilot studies demonstrated that there was no difference between phenylephrine dose-response curves over time (data not shown); however, after a single dose of methacholine the phenylephrine doseresponse curve was significantly blunted (EC so = 1.4 ± 0.08 vs 1.95 ± 0.17 fLmo1!L; p < 0.05). Therefore it was only at the end of each experiment that methacholine was used to confirm a functional endothelium. There was a concentration-dependent constriction in mesenteric arteries in response to the cumulative addi-
10-8
10-5
Phenylephrine (mil) Fig. 2. Concentration-response curves to phenylephrine for mesenteric arteries from pregnant rats in absence (solid triangles) or presence (open triangles) of meclofenamate and from nonpregnant rats in absence (solid circles) or presence (open circles) ofmeclofenamate. Tension is expressed as a percentage of maximum phenylephrine response. Data represent mean ± SE for six experiments in each group.
tion of phenylephrine. Fig. 1 illustrates the average phenylephrine dose-response curves for arteries from pregnant and nonpregnant rats. The arteries from the pregnant rats were less sensitive to phenylephrine than were the arteries from the nonpregnant rats. This was indicated by the shift to the right in the dose-response curve of the arteries from the pregnant rats. As shown in Table I, the EC so was 2.26 ± 0.19 fLmo1!L for the arteries of the pregnant rats versus 1.48 ± 0.09 fLmo1!L (p < 0.01) for the nonpregnant rats. The phenylephrine-generated maximal tension was not different in the arteries from the pregnant rats (2.9 ± 0.18 mN/mm) versus the nonpregnant rats (3.1 ± 0.12 nM/mm). The average dose-response curves of arteries to phenylephrine in the absence or presence of meclofenamate is shown in Fig. 2. Meclofenamate had no effect on the phenylephrine response in the arteries from either the pregnant or the nonpregnant animals. The EC so response to phenylephrine for the arteries of the pregnant rats was 2.26 ± 0.19 fLmo1!L versus 2.06 ± 0.14 fLmo1!L in the presence of meclofenamate (Table I). For the arteries from the nonpregnant rats the EC so response to phenylephrine was 1.48 ± 0.09 fLmo1!L versus 1.55 ± 0.15 fLmo1!L in the presence of meclofenamate (Table I). The similar dose ratios of approximately 1 for both groups of animals further illustrates that there was no meclofenamate treatment effect for arteries in either the pregnant or nonpreg-
1694 Davidge and Mclaughlin
December 1992 Am J Obstet Gynecol
Table I. Sensitivity and some ratios for phenylephrine in absence or presence of meclofenamate, LNAME, or endothelium removal in mesenteric arteries of nonpregnant and pregnant rats Pregnant Treatment Phenylephrine alone Phenylephrine with meclofenamate Phenylephrine with LNAME Phenylephrine with endothelium removal Values are *p < 0.01, tp < 0.01, tp < 0.01,
EC50 (,.Lmol!L) 2.26 2.06 1.11 1.12
± ± ± ±
0.19* 0.14* 0.11* t 0.20* t
I
Nonpregnant Dose ratio 1.13 ± 0.33 2.04 ± 0.13 1.96 ± 0.32
EC50 (p,mol!L) 1.48 1.55 0.72 0.74
± ± ± ±
0.09 0.15 0.09t 0.04t
I
Dose ratio 0.97 ± 0.12 2.13 ± 0.24 2.02 ± 0.11
mean ± SE. pregnant versus nonpregnant. significant increase in phenylephrine sensitivity with addition of LNAME. significant increase in phenylephrine sensitivity with endothelium removal.
nant rats (Table I). The phenylephrine-generated maximal tension did not change in the presence of meclofenamate (data not shown). The average dose-response curves of arteries to phenylephrine in the absence or presence of LNAME is shown in Fig. 3. LNAME shifted the phenylephrine dose-response curve to the left in both groups. This inhibition of EDRF increased sensitivity twofold in arteries from both the pregnant and nonpregnant rats, as evidenced by their equivalent dose ratios (Table I). The EC 50 response to phenylephrine for the arteries of the pregnant rats was 2.26 ± 0.19 versus 1.11 ± 0.11 I-Lmol/L (P < 0.01) in the presence of LNAME (Table I). For the arteries from the nonpregnant rats the EC 50 response to phenylephrine was 1.55 ± .15 versus 0.72 ± 0.09 I-Lmol/L (P < 0.01) in the presence of LNAME (Table I). Because the increased phenylephrine sensitivity in the presence of LNAME was similar between the two groups, the arteries from the pregnant rats remained less sensitive to phenylephrine than the arteries from the nonpregnant rats. Mter LNAME, the EC 50 was 1.11 ± 0.11 I-Lmol/L for the arteries of the pregnant rats versus 0.72 ± 0.09 I-Lmol/L (P < 0.01) for the nonpregnant rats. The phenylephrine-generated maximal tension did not change in the presence of LNAME (data not shown). Fig. 4 is a depiction of the average dose-response curves to phenylephrine in arteries with and without endothelium. Similar to the results of EDRF inhibition, endothelial cell removal caused an equivalent increase in phenylephrine sensitivity in the arteries of both the pregnant and nonpregnant rats. The EC 50 response to phenylephrine for the arteries of the pregnant rats was 2.26 ± 0.19 versus 1.12 ± 0.20 I-Lmol/L (P < 0.01) in the absence of endothelium (Table I). For the arteries from the nonpregnant rats the EC 50 response to phenylephrine was 1.48 ± 0.09 versus 0.74 ± 0.04 I-Lmol/L (P < 0.01) in the absence of endothelium (Table I). Because this increase in phenylephrine sensitivity was similar between the two groups, the arteries from the
pregnant rats remained less sensitive to phenylephrine than did the arteries from the nonpregnant rats. Mter endothelium removal the EC 50 was 1.12 ± 0.20 I-Lmol/L for the arteries ofthe pregnant rats versus 0.74 ± 0.04 I-Lmol/L (P < 0.01) for the nonpregnant rats. The similar dose ratios of approximately 2 for both groups of animals further illustrates that there was a similar effect from endothelium removal for arteries in both the pregnant and nonpregnant rats (Table I). The phenylephrine-generated maximal tension did not change with endothelium removal (data not shown). Arteries preconstricted with phenylephrine to their EC 50 values showed a concentration-dependent relaxation to methacholine (Fig. 5). Arteries from pregnant rats were significantly more sensitive to relaxation by methacholine than were the arteries from the nonpregnant rats (0.030 vs 0.049 fLmol/L, p < 0.01). However, there was no significant difference in the response of arteries from pregnant and nonpregnant rats to relaxation by sodium nitroprusside (0.10 vs 0.11 nmol/L). Relaxation with methacholine but not sodium nitroprusside was inhibited by endothelium removal or by LNAME (data not shown), suggesting that the effect of methacholine but not sodium nitroprusside was dependent on EDRF or a functional endothelium.
Comment The purpose of this study was to test if the reduction in adrenergic sensitivity that occurs during prengancy is modulated in part by endogenous vasorelaxants of the vessel wall. The major findings of the study are as follows: (1) There was a decrease in sensitivity to phenylephrine in mesenteric arteries from late-pregnant rats; (2) a cyclooxygenase inhibitor, meclofenamate, had no effect on this phenylephrine response in the arteries from either pregnant or nonpregnant rats; (3) an EDRF inhibitor, LNAME, increased the sensitivity to phenylephrine in arteries from both pregnant and nonpregnant rats, and this increase in sensitivity was similar between the two groups; (4) the effects of LNAME were
Endogenous modulation of vasoconstriction
Volume 167 Number 6
100 tI II
C
0
0. tI
100 tI II
c
80
II tI
D::
D::
E ~ E
60
....0 .... i·
e
tI Q,
60
'x0
'x0
:2
80
0 0.
II
E ~ E
1695
:2
....0 ....c
40
40
tI 0
L.
20
20
tI Q,
0
0 10-8
10-8
10-5
Phenylephrine (mil)
Phenylephrine (mil) Fig. 3. Concentration-response curves to phenylephrine for mesenteric arteries from pregnant rats in absence (solid triangles) or presence (open triangles) ofLNAME and from nonpregnant rats in absence (solid circles) or presence (open circles) of LNAME. Tension is expressed as a percentage of maximum phenylephrine response. Data represent mean ± SE for six experiments in each group.
mimicked by endothelium removal; (5) preconstricted arteries from pregnant rats were more sensitive to methacholine but not to sodium nitroprusside relaxation than were arteries from the nonpregnant rats. The decrease in sensitivity to adrenergic vasoconstrictors during pregnancy has been reported previously in both in vivo lO . 12 and in vitro. 4 . 9 However, most of the in vitro work was conducted in the aortas or the uterine vasculature. We have chosen to study resistance-sized arteries from the splanchnic circulation, because this a major component of total peripheral vascular resistance and therefore an important determinant of maternal systemic blood pressure. Previous work from our laboratory has shown a decreased sensitivity to norepinephrine-induced adrenergic vasoconstriction during pregnancy in resistance-sized mesenteric arteries. 13 Our results from this study extend and confirm these observations. A potential role for vasodilatory prostaglandins as mediators of the blunted response to vasoconstrictors during pregnancy was proposed a number of years ago. The increase of 6-keto-prostaglandin FIn (the stable metabolite of prostacydin) in plasma and urine during pregnancyl7 suggested that this particular vasodilatory prostaglandin was one such mediator of vasoreactivity. In addition, inhibition of cyclooxygenase with indomethacin or aspirin caused an increase in pressor responsiveness to angiotensin II in pregnant women. 18
10-5
Fig. 4. Concentration-response curves to phenylephrine for mesenteric arteries from pregnant rats in presence (solid triangles) or absence (open triangles) of endothelium and from nonpregnant rats in presence (solid circles) or absence (open circles) of endothelium. Tension is expressed as a percentage of maximum phenylephrine response. Data represent mean ± SE for six experiments in each group.
0 c
0
:;::;
25
0
)(
0
q; a:::
....c
., .,a..u
50
L.
75
~~ \,
.~
100
10-7
10-6
Methacholine (mil) Fig. 5. Concentration-response curves to methacholine for mesenteric arteries from pregnant (n = 6, solid triangles) and nonpregnant (n = 6, solid circles) rats. Responses are expressed as a percentage of relaxation from phenylephrine preconstricted levels. Data represent mean ± SE.
Similar data from whole-animal models have further supported this role of vasodilatory prostaglandins as mediators of the blunted pressor response during pregnancy.ID However, more recent studies have resulted in
1696 Davidge and Mclaughlin
conflicting data. In chronically instrumented. trained. conscious animals treatment with prostaglandin inhibitors was not able to reverse the blunted pressor response to adrenergic vasoconstriction observed during pregnancy.4. 11 In fact. with isolated vessels there is actually very little direct evidence that vasodilatory prostaglandins are responsible for decreased sensitivity to adrenergic vasoconstriction. In aortas from guinea pigs and rats. prostaglandin inhibitors did not change the contractile response to adrenergic agents in either pregnant or nonpregnant animals. 4. 9 The results from our study with resistance-sized mesenteric arteries are in agreement with the isolated aortic ring studies. The prostaglandin products of the cyclooxygenase pathway do not appear to modulate acutely the in vitro vascular reactivity of arteries from either the pregnant or nonpregnant rats. However. our study design did not test the possibility that alterations in prostaglandin metabolism during pregnancy may have chronic effects on vascular smooth muscle function that are not apparent with acute cyclooxygenase inhibition. Another important vasorelaxant of the vessel wall is EDRF. which may have a potential role as a mediator for the blunted vasoconstrictor response observed during pregnancy. In both in vivo and in vitro systems. EDRF is known to play an important role in controlling blood vessel reactivity. It has been reported that inhibition of EDRF results in an increased sensitivity to adrenergic vasoconstriction. 19. 20 However, the role of EDRF modulation of a blunted vasoconstrictor response during pregnancy remains controversial. In conscious. instrumented nonpregnant and pregnant rats. a bolus infusion of Nw-nitro-L-arginine (an inhibitor of EDRF synthase) produced similar increases in blood pressure. 12 These data suggested that pregnancy does not result in a greater role for EDRF in the maintenance of basal blood pressure. Similar results have been reported in vitro. Parent et aI.' reported that the endothelium of the superior mesenteric artery of nonpregnant and pregnant rats similarly influenced the response to adrenergic vasoconstriction. It has also been shown that neither EDRF inhibition nor endothelium removal results in a greater adrenergic vasoconstrictor response in the aorta of the pregnant rat. 5 . 9The results from our study with resistance-sized mesenteric arteries are in agreement with the aortic ring studies in that inhibition ofEDRF or endothelial cell removal had similar increases in phenylephrine sensitivity in arteries from the pregnant and nonpregnant rats. This suggests that there is a similar endothelium-dependent modulation to a vasoconstrictor response for arteries from both the pregnant and nonpregnant rats. Contrary to the data with arteries from rats, endothelial removal or inhibition of EDRF augmented the norepinephrine
December 1992 Am J Obstet Gynecol
sensitivity more in isolated aortas and uterine arteries from pregnant guinea pigs than it did from the nonpregnant animals. 6 • 8 Because there are elevated plasma and urinary levels of cyclic guanosine 5' -monophosphate (the second messenger involved in the action ofEDRF) during gestation in the rat. 21 we had expected that the blunted response to phenylephrine during pregnancy would be caused in part by augmented EDRF activity. Although our results do not support this. it is possible that our findings are a function of the study design. If the elevated cyclic guanosine 5' -monophosphate levels indicate enhanced EDRF synthesis during pregnancy. the isolated arteries from the pregnant rats may have become L:-arginine deficient over the course of the study. Inhibition of nitrate synthase would then underestimate the contributory role of EDRF to the phenylephrine response. L-Arginine deficiency in the arteries from the pregnant rats is unlikely in our experiments. because the vessels maintained their responses to repetitive methacholine relaxation. It would be interesting to determine if pregnancy induces an isoform of nitrate synthase in the endothelium or smooth muscle that differs from the constitutive enzyme. Finally, the effect of chronic EDRF inhibition (or combined blockade of EDRF and cyclooxygenase) in vitro was not examined. Such inhibition may be required to reverse possible longer term effects of EDRF on the vascular smooth muscle. For example. nitric oxide has recently been shown to react with membrane-bound thiol groups on the N-methyl-D-aspartate (NMDA) subtype of glutamate receptor, resulting in a persistent blockade of NMDA responses!· There may be a role for EDRF in modulating vasoconstrictor responses in vivo, one that is not apparent when examined in vitro. These in vitro studies are conducted under conditions of no flow; however, flowinduced shear stress in vivo can cause production and release of ED RF from the endothelium.23 Therefore, in addition to investigating the gestational effect on basal endothelial influence, we were also interested in the gestational effect on stimulated endothelium-dependent relaxation. The effect of gestation on endothelium-dependent vasorelaxation has been investigated by looking at stimulated EDRF-mediated responses in preconstricted arterial segments or by measuring the accumulation of cyclic guanosine 5'-monophosphate in isolated arterial preparations. Again the data are conflicting as to the role EDRF has in modulating vasoactivity during pregnancy. In isolated aortas from pregnant rats cyclic guanosine 5' -monophosphate production in response to the addition of methacholine was increased compared with that in nonpregnant rats. 24 Furthermore, preconstricted uterine arteries from pregnant guinea
Volume 167 Number 6
pigs are more sensitive to endothelium-dependent relaxations, as were renal interlobar arteries from pregnant rats. 25 • 26 However, there was no difference in the endothelium-dependent relaxation of preconstricted aortas from virgin and pregnant rats when the constricting agent was norepinephrine or phenylephrine,s. 12 although there was a difference if the arteries were preconstricted with the thromboxane A2 sympathomimetic U46619. 5 In our laboratory we have observed that the potential for endothelium-dependent relaxation becomes greater with decreasing artery size, on the basis of studies with the use of the aorta and the superior and first-, secondo, and third-order branches of the mesenteric arteries (unpublished observations). Therefore the contribution that the endothelium makes in modulating vasoactivity may be more important in the resistance-sized arteries than that of the conduit arteries such as the aorta. The results from our study demonstrated that preconstricted mesenteric arteries from pregnant rats are more sensitive to relaxation by methacholine than are arteries of the nonpregnant rats. It is likely that this relaxation induced by methacholine was EDRF dependent because the relaxation was inhibited by endothelium removal or by LNAME. Furthermore, the enhanced relaxation to methacholine in arteries from the pregnant rats is most likely due to a greater release of EDRF. This can be proposed because there was no difference in relaxation response to the endothelium-independent agonist sodium nitroprusside from arteries of the pregnant and nonpregnant rats. Whether other nonmuscarinic, endothelium-dependent agents produce a similar potentiated response in these arteries from pregnant rats is unknown. In conclusion, these studies demonstrated that the decreased sensitivity to adrenergic agonists during pregnancy in the mesenteric vasculature is not modulated acutely by prostaglandin products of the cyclooxygenase pathway or by EDRF. However, the potential for endothelium-dependent relaxation is augmented in pregnancy. Because our data indicate that the vasoconstrictor response to phenylephrine is reduced by late gestation, this may reflect a generalized reduction in the smooth muscle response to vasoconstrictors. It is possible that a reduced vascular smooth muscle response to adrenergic stimulation and an increased capacity to release EDRF interact to contribute to the systemic vasodilatation observed during pregnancy. REFERENCES I. de Swiet M. Cardiovascular system. In: Hytten F, Chamberlain G, eds. Clinical physiology in obstetrics. Boston: Blackwell Scientific, 1980:3-42. 2. Gant NF, Daley GL, Chand S, Whalley P], MacDonald PC. A study of angiotensin II pressor response throughout primigravid pregnancy.] Clin Invest 1973;52:2682-9.
Endogenous modulation of vasoconstriction
1697
3. Vanhoutte PM, Rubanyi GM, Miller VM, Houston DS. Modulation of vascular smooth muscle contraction by the endothelium. Ann Rev Physiol 1986;48:307-20. 4. Harrison GL, Moore LG. Blunted vasoreactivity in pregnant guinea pigs is not restored by meclofenamate. AM] OSSTET GYNECOL 1989; 160:258-64. 5. ]ansakul C, Boura ALA, King RG. Effects of endothelial cell removal on constrictor and dilator responses of aortae of pregnant rats.] Auton Pharmacol 1989;9:93-101. 6. Moore LG. Endothelial contribution to decreased contractility in isolated aortic rings from pregnant compared to nonpregnant animals [Abstract). FASEB] 1989;3:A259. 7. Parent A, Schiffrin EL, St-Louis]. Role of the endothelium in adrenergic responses of mesenteric artery rings of pregnant rats. AM] OSSTET GYNECOL 1990;163:229-34. 8. Weiner C, Liu KZ, Thompson L, Herrig ], Chestnut D. Effect of pregnancy on endothelium and smooth muscle: their role in reduced adrenergic sensitivity. Am] Physiol 1991 ;261:H 1275-83. 9. St-Louis], Sicotte B. Prostaglandin- or endothelium-mediated vasodilation is not involved in the blunted responses of blood vessels to vasoconstrictors in pregnant rats. AM] OSSTET GYNECOL 1992; 166:684-92. 10. Pallor MS. Mechanism of decreased responsiveness to angiotensin II, NE and vasopressin in pregnant rats. Am] Physiol 1984;247:HI00-8. 11. Conrad KP, Colpoys MC. Evidence against the hypothesis that prostaglandins are the vasodepressor agents of pregnancy: serial studies in chronically instrumented, conscious rats.] Clin Invest 1986;77;236-45. 12. Umans ]G, Lindheimer MD, Barron WM. Pressor effect of endothelium-derived relaxing factor inhibition in conscious virgin and gravid rats. Am ] Physiol 1990;259: F293-6. 13. Crandall ME, Keve TM, McLaughlin MK. Characterization of norepinephrine sensitivity in the maternal splanchnic circulation during pregnancy. AM ] OBSTET GYNECOL 1990; 162: 1296-30 1. 14. McLaughlin MK, Keve TM. Pregnancy-induced changes in resistance blood vessels. AM ] OBSTET GYNECOL 1986; 155: 1296-9. 15. Osol G, Cipolla M, Knutson S. A new method for mechanically denuding the endothelium of small (50-150 J.l.m) arteries with a human hair. Blood Vessels 1989;26:320-4. 16. Hill A. The combination of hemoglobin with oxygen and carbon monoxide. Biochem] 1913;7:471-80. 17. Goodman RP, Killam AP, Brash AR, Branch RA. Prostacyclin production during pregnancy: a comparison of production during normal pregnancy and pregnancy complicated by hypertension. AM] OBSTET GYNECOL 1982;142: 817-22. 18. Everett RB, Worley R], MacDonald PC, Gant NF. Effect of prostaglandin synthetase inhibitors on pressor response to angiotensin II in human pregnancy. ] Clin Endocrinol Metab 1978;46:1007-10. 19. Martin W, Furchgott RF, Villani GM, ]othianandan D. Depression of contractile responses in rat aorta by spontaneously released endothelium-derived relaxing factor. ] Pharmacol Exp Ther 1986;237:529-38. 20. Fukuda S, Matsumoto M, Nishimura N, et al. Endothelial modulation of norepinephrine-induced constriction of rat aorta at normal and high CO 2 tensions. Am ] Physiol 1990;258:H 1049-54. 21. Conrad KP, Vernier KA. Plasma level, urinary excretion, and metabolic production of cGMP during gestation in rats. Am] Physiol 1989;257:R847-53. 22. Lei SZ, Pan Z-H, Aggarwal SK, et al. Effect of nitric oxide production on the redox modulatory site of the NMDA receptor-channel complex. Neuron 1992;8:1087-99. 23. Furchgott RF, Vanhoutte PM. Endothelium-derived relaxing and contracting factors. FASEB] 1989;3:2007-18.
Davidge and Mclaughlin
24. Whittemore SL, Conrad KP. Stimulation of cyclic guanosine 5' -monophosphate (cGMP) production by methylcholine differs in aortae from pregnant and virgin rats [Abstract]. Physiologist 1989;32: 186. 25. Weiner CP, Martinez E, Zhu LK, Ghodsi A, Chestnut DH. In-vitro release of endothelium-derived relaxing factor by acetylcholine is increased during the guinea pig pregnancy. AM] OSSTET GYNECOL 1989;161:1599-605.
December 1992 Am J Obstet Gynecol
26. Griggs KC, McLaughlin MK. Endothelial modulation of vasoconstrictor response of isolated renal interlobar arteries from pregnant and nonpregnant rats [Abstract]. In: Proceedings of the thirty-eighth annual meeting of the Society for Gynecologic Investigation, San Antonio, Texas, March 20-23, 1991. San Antonio: Society for Gynecologic Investigation, 1991.
A unique hypertonic response to hypotonic infusion in the pregnant ewe David R. Powers, MD,· and Robert A. Brace, PhDb
San Diego, California OBJECTIVE: The purpose of this study was to compare the responses of the maternal ewe to intravenous volume expansion with either sufficient lactated Ringer's solution to elevate maternal venous pressure or sufficient hypotonic fluid to reduce blood osmolality. STUDY DESIGN: Chronically catheterized pregnant sheep were intravenously infused over 4 hours with either commercial lactated Ringer's solution (5.55 ± 0.50 Uhr, 255 mOsm/kg, mildly hypotonic) or diluted Ringer's solution (2.04 ± 0.27 Uhr, 150 mOsm/kg, markedly hypotonic). Data were statistically analyzed with two- and three-factor analyses of variance and bivariate regression analysis. RESULTS: During the mildly hypotonic infusion (n = 8) the maternal blood osmolality changes were - 5.1 ± 1.2, + 2.7 ± 1.0 and + 6.8 ± 1.1 mOsm/kg at 1 and 4 hours of infusion and 1 hour after the infusion. In four of the eight animals in this group profuse diarrhea developed. During the markedly hypotonic infusion (n = 11) the maternal blood osmolality changes were - 9.9 ± 1.1, -15.9 ± 2.5, and -10.4 ± 2.2 mOsm/kg at 1 and 4 hours of infusion and 1 hour after the infUSion. Although urine osmolalities were significantly less than the osmolality of the infusate in both groups, only during the mildly hypotonic infusion was there a net loss of free water by the kidneys. The renal free water loss, the venous pressure increase, and the blood osmolality decrease were not significantly different whether diarrhea did or did not develop. CONCLUSION: The infusion of large volumes of mildly hypotonic Ringer's solution to the pregnant ewe produces a paradoxic increase in maternal plasma osmolality as a result of the excretion of large volumes of free water by the kidneys, and if the venous pressure is increased more than about 6 mm Hg with this infusion, diarrhea develops in the animals. (AM J OesTET GVNECOL 1992;167:1698-709.)
Key words: Pregnancy, osmolality, ewe, free water clearance, diarrhea Maternal volume status changes dramatically during human pregnancy. Blood volume increases by about
From the Division of Perinatal Medicine, Department of Reproductive Medicine, University of California at San Diego, a, b and the Division of Nephrology, Department of Internal Medicine, University of California at Irvine. a Supported in part by National Institutes of Health grants HD20295 and HD23724 from the National Institute of Child Health and Human Development. Presented at the Thirty-eighth Annual Meeting of the Society for Gynecologic Investigation, San Antonio, Texas, March 21-23,1991. Reprint requests: David R. Powers, MD, Department of Reproductive Medicine, University of California at San Diego, La Jolla, CA 92093-0802.
6/6/41716 1698
1.5 L, and the change in plasma volume accounts for the majority of the increase. I - 3 Furthermore, a failure of maternal plasma volume to expand has been associated with disease states such as preeclampsia,., 5 chronic hypertension,5. 6 intrauterine growth retardation, 7 and 0ligohydramnios. 8 If maternal volume status is indeed compromised in these conditions, the result could be a decrease in maternal hydrostatic pressure perfusing the placenta or an increase in osmotic pressure of maternal plasma, both potentially decreasing fluid movement to the fetus. In spite of recent studies suggesting that maternal volume expansion may be of benefit in the treatment of preeclampsia9 or oligohydramnios,1O little is currently known about the effects of fluid therapy on
Cincinnati, Ohio OBJECTIVE: We tested the hypothesis that during pregnancy the endothelium mediates the blunted response to adrenergic vasoconstriction. STUDY DESIGN: Mesenteric resistance arteries from late pregnant (n = 6) and age-matched virgin control (n = 6) Sprague-Dawley rats were studied in a myograph. RESULTS: Arteries from pregnant rats were 35% less sensitive to phenylephrine vasoconstriction than were those from nonpregnant rats (mean effective concentration that produced a 50% response 2.26 vs 1.48 Il-mol/L, pregnant vs nonpregnant, p < 0.01). Meclofenamate had no effect on the vasoconstrictor response in arteries from either group. Inhibition of endothelium-derived relaxing factor with W-nitro-L-arginine methyl ester or endothelial cell removal had a similar twofold increase in phenylephrine sensitivity in arteries from both the pregnant and nonpregnant rats (mean effective concentration that produced a 50% response 2.26 vs 1.11 Il-mol/L for pregnant rats and 1.48 vs 0.72 Il-mol/L for nonpregnant rats, p < 0.01). However, methacholine relaxation response was potentiated in pregnant versus nonpregnant rats (mean effective concentration that produced a 50% response 0.030 vs 0.049 Il-mol/L, p < 0.01). CONCLUSION: Although the potential for endothelium-dependent relaxation is augmented in mesenteric arteries of the pregnant rat, the decreased sensitivity to phenylephrine during pregnancy is not modulated acutely by endothelium-derived relaxing factor or by prostaglandin products of the cyclooxygenase pathway. (AM J OSSTET GYNECOL 1992;167:1691-8.)
Key words: Pregnancy, adrenergic vascular reactivity, resistance arteries, endothelium, Sprague-Dawley rat The maternal hemodynamics of normal pregnancy have been well described. This includes the decrease in peripheral vascular resistance that is associated with a decreased pressor responsiveness to vasoconstrictors. I This reduction in vascular reactivity is clinically important because it is not evident in women with pregnancyinduced hypertension or preeclampsia. 2 In spite of considerable research the mechanisms that could account for this decreased pressor responsiveness during pregnancy are not well understood. Specific alterations within the vascular wall are likely to contribute significantly to this gestational change in vascular reactivity. The endothelium produces and releases substances such as prostacyclin and endothelium-derived From the Perinatal Research Institute, Division of Neonatology, Department of Pediatrics, University of Cincinnati College of Medicine. Supported in part by United States Public Health Service grant No. 40130.
Presented at the Thirty-ninth Annual Meeting of the Society for Gynecologic Investigation, San Antonio, Texas, March 18-21, 1992. Reprint requests: Margaret K. McLaughlin, PhD, Children's Hospital Medical Center, TCHRF - Neonatology Division, Eiland and Bethesda Ave., Cincinnati, OH 45229-2899. 6/6/41903
relaxing factor (EDRF) that will modulate the constriction of vascular smooth muscle. 3 However, it remains controversial if these vasorelaxants are involved in the decreased vasoreactivity observed during pregnancy.<-12 One reason for the conflicting data may involve the use of conduit arteries from different vascular beds. Therefore we have chosen to study resistance-sized mesenteric arteries because this vascular bed is a major determinant of total peripheral vascular resistance. This study was designed to test the hypothesis that the reduction in adrenergic reactivity that occurs in resistance-sized mesenteric arteries during pergnancy13 is in part modulated by endogenous vasorelaxants of the vessel wall. In particular, we were interested in studying the differential influence of EDRF and products of the cyclooxygenase pathway in modulating the vasoconstrictor response to phenylephrine in arteries from pregnant and nonpregnant rats.
Methods General animal model. Ten-week-old female Sprague-Dawley rats were obtained from Harlan (Harlan, Ind.) and bred in our own colony. Each morning vaginal smears were examined microscopically, and the
1691
1692 Davidge and Mclaughlin
presence of sperm was considered day 1 of gestation (term 22 days). The experiments were performed during late (18 to 19 days) gestation and in age-matched virgin controls. The animals were housed in the facilities of the Department of Laboratory Animal Medical Services at the University of Cincinnati, which is accredited by the American Association for the Accreditation of Laboratory Animal Care. Purina 5001 Rat Chow and water were provided ad libitum. Vessel preparation. Rats were decapitated while they were under light anesthesia with methohexital sodium (50 mg/kg body weight). A section of the mesentery 5 to 10 cm distal to the pylorus was rapidly removed and placed in ice-cold N-[2-hydroxyethyl]piperazine-N' -[2ethanesulfonic acid] (HEPES)-buffered physiologic saline solution. One mesenteric artery averaging 250 fLm in diameter was dissected free from surrounding adipose tissue, cut into two 1.8 cm lengths, and threaded onto 16 fLm wires. These wires were fastened to two stainless steel blocks that were mounted in a specially designed myograph system. 14 One block was attached to a Kulite strain gauge force transducer (Kulite Semiconductor Products, Ridgefield, N.J.), and the other was connected to a displacement device. The blocks rested in 10 ml glass-jacketed organ baths with HEPES-physiologic saline solution, kept at 37° C. Each experiment had two baths running simultaneously. Resting length-tension curve. Mter being mounted, the arteries were stretched to about 0.2 nM/mm vessel length (1 mN = 102 mg) and allowed to equilibrate for 1 hour in HEPES-physiologic saline solution buffer. The arteries were then given a conditioning stretch of about 0.6 mN. To generate the resting curve, each artery was stretched in six incremental steps and was held for 20 seconds at each step. The force at each stretch was read at the end of 20 seconds, and the displacement was measured. We have termed this the resting length-tension relationship. The arterial circumference that was used to perform the dose-response curves was obtained by using the Law of LaPlace. With this equation LIOO is calculated from the exponential curve fit of tension versus circumference. LIOO is defined as the circumference the vessel would have at a transmural pressure of 100 mm Hg. We have found from our previous studies that a doseresponse curve obtained at 0.8 LIOO is a point that provides maximum active force generation with minimum passive tension in arteries from both nonpregnant and pregnant female rats. Solutions and drugs. The HEPES-physiologic saline solution used in these experiments contained sodium chloride 142 mmol!L, potassium chloride 4.7 mmol!L, magnesium sulfate 1.17 mmol!L, calcium chloride 1.56 mmol/L, potassium phosphate 1.18 mmol!L, HEPES 10 mmol/L, and glucose 5.5 mmol!L. The HEPES-physi-
December 1992 Am J Obstet Gynecol
ologic saline solution was maintained at a pH of 7.4. Stock solutions of phenylephrine (L-phenylephrine hydrochloride), methacholine (acetyl-l3-methylcholine chloride), sodium nitroprusside (all from Sigma, St. Louis), and meclofenamate (Warner Lambert, Ann Arbor) were prepared in HE PES-physiologic saline solution at a concentration of 10 mmol!L for each experiment. A stock solution of 10 mmol!L of Nw-nitro-Larginine methol ester (LNAME, Sigma) was prepared in water. Appropriate dilutions of all stocks were obtained with HEPES-physiologic saline solution. Experimental design. Vasoreactivity was measured with the aI-adrenergic agonist phenylephrine. To determine the role of endogenous vasorelaxants in modulating the vasoconstrictor response to phenylephrine, meclofenamate was administered to block cyclooxygenase conversion of arachidonic acid and LNAME was used to block the conversion of L-arginine to EDRF. The order of drug administration and the time between drug exposure for these experiments were determined in a preliminary set of experiments designed to test for tachyphylaxis and drug interaction. Cumulative doses of phenylephrine (0.3 to 10 fLmol! L) were administered 30 minutes after the completion of a resting-length tension curve. Mter completion of the dose-response curve, a 30-minute recovery period was allowed, during which the baths were frequently changed with fresh HEPES-physiologic saline solution. By means of two separate baths, one segment of the divided artery from each rat had the phenylephrine dose-response curve repeated in the presence of meclofenamate (1 fLmol!L), whereas the other segment had the phenylephrine dose-response curve repeated in the presence ofLNAME (0.25 mmol!L). Meclofenamate was added 30 minutes before and LNAME 10 minutes before the repeat phenylephrine response curve was generated. At the end of each experiment the arteries were preconstricted with phenylephrine to 50% of their maximal response, and a single dose of methacholine (1 fLmol!L) was administered to prove that a functional endothelium existed for these experiments. A second set of arteries from the same rats were denuded of endothelium before the administration of cumulative doses of phenylephrine (0.3 to 10 fLmol!L). Endothelium removal was done mechanically with a human hair threaded through the lumen of the artery and rubbed back and forth.15 Confirmation of complete endothelium removal was done pharmacologically with a single dose of 1 fLmol!L methacholine at the end of the experiment. Previous experiments with silver staining and scanning electron microscopy had shown that pharmacologic confirmation of endothelial cell removal is appropriate for these arteries (unpublished observation). In a separate protocol with mesenteric arteries from a
Endogenous modulation of vasoconstriction
Volume 167 Number 6
100
• II
C
0
a.
100
•c II
80
0
a.
•
II
0::
E E
0::
60
~
0
....c
•f:! •0..
E E
60
:::I
40
::J
'xg
:::I
80
•
II
::J
1693
'xg
40
~
0
....c
•f:! •
20
0..
20
0
0
10- 8
10-5
Phenylephrine (mil) Fig. 1. Concentration-response curves to phenylephrine for mesenteric arteries from pregnant (n = 6, solid triangles) and nonpregnant (n = 6, solid circles) rats. Tension is expressed as a percentage of maximum phenylephrine response. Data represent mean ± SE.
different group of rats, a cumulative dose-resonse curve to the endothelium-dependent vasorelaxant methacholine (10 nmo1!L to 1 fLmo1!L) and the endotheliumindependent vasorelaxant sodium nitroprusside (1 nmo1!L to 0.1 fLmo1!L) was obtained after preconstricting the arteries with phenylephrine to 50% of the maximum response. Data analysis. The data from the dose-response curves were fitted to the Hill equation,16 from which a straight line was generated by linear least-squares regression analysis. The mean effective concentration that produced a 50% response (EC so ) was determined from this line and expressed as the geometric mean ± SE. Dose ratios were determined as the ratio of the EC so before treatment to that after treatment. Comparisons between groups and treatments were done with a twoway analysis of variance with repeated measures. A value of p < 0.05 was considered significant.
Results The pilot studies demonstrated that there was no difference between phenylephrine dose-response curves over time (data not shown); however, after a single dose of methacholine the phenylephrine doseresponse curve was significantly blunted (EC so = 1.4 ± 0.08 vs 1.95 ± 0.17 fLmo1!L; p < 0.05). Therefore it was only at the end of each experiment that methacholine was used to confirm a functional endothelium. There was a concentration-dependent constriction in mesenteric arteries in response to the cumulative addi-
10-8
10-5
Phenylephrine (mil) Fig. 2. Concentration-response curves to phenylephrine for mesenteric arteries from pregnant rats in absence (solid triangles) or presence (open triangles) of meclofenamate and from nonpregnant rats in absence (solid circles) or presence (open circles) ofmeclofenamate. Tension is expressed as a percentage of maximum phenylephrine response. Data represent mean ± SE for six experiments in each group.
tion of phenylephrine. Fig. 1 illustrates the average phenylephrine dose-response curves for arteries from pregnant and nonpregnant rats. The arteries from the pregnant rats were less sensitive to phenylephrine than were the arteries from the nonpregnant rats. This was indicated by the shift to the right in the dose-response curve of the arteries from the pregnant rats. As shown in Table I, the EC so was 2.26 ± 0.19 fLmo1!L for the arteries of the pregnant rats versus 1.48 ± 0.09 fLmo1!L (p < 0.01) for the nonpregnant rats. The phenylephrine-generated maximal tension was not different in the arteries from the pregnant rats (2.9 ± 0.18 mN/mm) versus the nonpregnant rats (3.1 ± 0.12 nM/mm). The average dose-response curves of arteries to phenylephrine in the absence or presence of meclofenamate is shown in Fig. 2. Meclofenamate had no effect on the phenylephrine response in the arteries from either the pregnant or the nonpregnant animals. The EC so response to phenylephrine for the arteries of the pregnant rats was 2.26 ± 0.19 fLmo1!L versus 2.06 ± 0.14 fLmo1!L in the presence of meclofenamate (Table I). For the arteries from the nonpregnant rats the EC so response to phenylephrine was 1.48 ± 0.09 fLmo1!L versus 1.55 ± 0.15 fLmo1!L in the presence of meclofenamate (Table I). The similar dose ratios of approximately 1 for both groups of animals further illustrates that there was no meclofenamate treatment effect for arteries in either the pregnant or nonpreg-
1694 Davidge and Mclaughlin
December 1992 Am J Obstet Gynecol
Table I. Sensitivity and some ratios for phenylephrine in absence or presence of meclofenamate, LNAME, or endothelium removal in mesenteric arteries of nonpregnant and pregnant rats Pregnant Treatment Phenylephrine alone Phenylephrine with meclofenamate Phenylephrine with LNAME Phenylephrine with endothelium removal Values are *p < 0.01, tp < 0.01, tp < 0.01,
EC50 (,.Lmol!L) 2.26 2.06 1.11 1.12
± ± ± ±
0.19* 0.14* 0.11* t 0.20* t
I
Nonpregnant Dose ratio 1.13 ± 0.33 2.04 ± 0.13 1.96 ± 0.32
EC50 (p,mol!L) 1.48 1.55 0.72 0.74
± ± ± ±
0.09 0.15 0.09t 0.04t
I
Dose ratio 0.97 ± 0.12 2.13 ± 0.24 2.02 ± 0.11
mean ± SE. pregnant versus nonpregnant. significant increase in phenylephrine sensitivity with addition of LNAME. significant increase in phenylephrine sensitivity with endothelium removal.
nant rats (Table I). The phenylephrine-generated maximal tension did not change in the presence of meclofenamate (data not shown). The average dose-response curves of arteries to phenylephrine in the absence or presence of LNAME is shown in Fig. 3. LNAME shifted the phenylephrine dose-response curve to the left in both groups. This inhibition of EDRF increased sensitivity twofold in arteries from both the pregnant and nonpregnant rats, as evidenced by their equivalent dose ratios (Table I). The EC 50 response to phenylephrine for the arteries of the pregnant rats was 2.26 ± 0.19 versus 1.11 ± 0.11 I-Lmol/L (P < 0.01) in the presence of LNAME (Table I). For the arteries from the nonpregnant rats the EC 50 response to phenylephrine was 1.55 ± .15 versus 0.72 ± 0.09 I-Lmol/L (P < 0.01) in the presence of LNAME (Table I). Because the increased phenylephrine sensitivity in the presence of LNAME was similar between the two groups, the arteries from the pregnant rats remained less sensitive to phenylephrine than the arteries from the nonpregnant rats. Mter LNAME, the EC 50 was 1.11 ± 0.11 I-Lmol/L for the arteries of the pregnant rats versus 0.72 ± 0.09 I-Lmol/L (P < 0.01) for the nonpregnant rats. The phenylephrine-generated maximal tension did not change in the presence of LNAME (data not shown). Fig. 4 is a depiction of the average dose-response curves to phenylephrine in arteries with and without endothelium. Similar to the results of EDRF inhibition, endothelial cell removal caused an equivalent increase in phenylephrine sensitivity in the arteries of both the pregnant and nonpregnant rats. The EC 50 response to phenylephrine for the arteries of the pregnant rats was 2.26 ± 0.19 versus 1.12 ± 0.20 I-Lmol/L (P < 0.01) in the absence of endothelium (Table I). For the arteries from the nonpregnant rats the EC 50 response to phenylephrine was 1.48 ± 0.09 versus 0.74 ± 0.04 I-Lmol/L (P < 0.01) in the absence of endothelium (Table I). Because this increase in phenylephrine sensitivity was similar between the two groups, the arteries from the
pregnant rats remained less sensitive to phenylephrine than did the arteries from the nonpregnant rats. Mter endothelium removal the EC 50 was 1.12 ± 0.20 I-Lmol/L for the arteries ofthe pregnant rats versus 0.74 ± 0.04 I-Lmol/L (P < 0.01) for the nonpregnant rats. The similar dose ratios of approximately 2 for both groups of animals further illustrates that there was a similar effect from endothelium removal for arteries in both the pregnant and nonpregnant rats (Table I). The phenylephrine-generated maximal tension did not change with endothelium removal (data not shown). Arteries preconstricted with phenylephrine to their EC 50 values showed a concentration-dependent relaxation to methacholine (Fig. 5). Arteries from pregnant rats were significantly more sensitive to relaxation by methacholine than were the arteries from the nonpregnant rats (0.030 vs 0.049 fLmol/L, p < 0.01). However, there was no significant difference in the response of arteries from pregnant and nonpregnant rats to relaxation by sodium nitroprusside (0.10 vs 0.11 nmol/L). Relaxation with methacholine but not sodium nitroprusside was inhibited by endothelium removal or by LNAME (data not shown), suggesting that the effect of methacholine but not sodium nitroprusside was dependent on EDRF or a functional endothelium.
Comment The purpose of this study was to test if the reduction in adrenergic sensitivity that occurs during prengancy is modulated in part by endogenous vasorelaxants of the vessel wall. The major findings of the study are as follows: (1) There was a decrease in sensitivity to phenylephrine in mesenteric arteries from late-pregnant rats; (2) a cyclooxygenase inhibitor, meclofenamate, had no effect on this phenylephrine response in the arteries from either pregnant or nonpregnant rats; (3) an EDRF inhibitor, LNAME, increased the sensitivity to phenylephrine in arteries from both pregnant and nonpregnant rats, and this increase in sensitivity was similar between the two groups; (4) the effects of LNAME were
Endogenous modulation of vasoconstriction
Volume 167 Number 6
100 tI II
C
0
0. tI
100 tI II
c
80
II tI
D::
D::
E ~ E
60
....0 .... i·
e
tI Q,
60
'x0
'x0
:2
80
0 0.
II
E ~ E
1695
:2
....0 ....c
40
40
tI 0
L.
20
20
tI Q,
0
0 10-8
10-8
10-5
Phenylephrine (mil)
Phenylephrine (mil) Fig. 3. Concentration-response curves to phenylephrine for mesenteric arteries from pregnant rats in absence (solid triangles) or presence (open triangles) ofLNAME and from nonpregnant rats in absence (solid circles) or presence (open circles) of LNAME. Tension is expressed as a percentage of maximum phenylephrine response. Data represent mean ± SE for six experiments in each group.
mimicked by endothelium removal; (5) preconstricted arteries from pregnant rats were more sensitive to methacholine but not to sodium nitroprusside relaxation than were arteries from the nonpregnant rats. The decrease in sensitivity to adrenergic vasoconstrictors during pregnancy has been reported previously in both in vivo lO . 12 and in vitro. 4 . 9 However, most of the in vitro work was conducted in the aortas or the uterine vasculature. We have chosen to study resistance-sized arteries from the splanchnic circulation, because this a major component of total peripheral vascular resistance and therefore an important determinant of maternal systemic blood pressure. Previous work from our laboratory has shown a decreased sensitivity to norepinephrine-induced adrenergic vasoconstriction during pregnancy in resistance-sized mesenteric arteries. 13 Our results from this study extend and confirm these observations. A potential role for vasodilatory prostaglandins as mediators of the blunted response to vasoconstrictors during pregnancy was proposed a number of years ago. The increase of 6-keto-prostaglandin FIn (the stable metabolite of prostacydin) in plasma and urine during pregnancyl7 suggested that this particular vasodilatory prostaglandin was one such mediator of vasoreactivity. In addition, inhibition of cyclooxygenase with indomethacin or aspirin caused an increase in pressor responsiveness to angiotensin II in pregnant women. 18
10-5
Fig. 4. Concentration-response curves to phenylephrine for mesenteric arteries from pregnant rats in presence (solid triangles) or absence (open triangles) of endothelium and from nonpregnant rats in presence (solid circles) or absence (open circles) of endothelium. Tension is expressed as a percentage of maximum phenylephrine response. Data represent mean ± SE for six experiments in each group.
0 c
0
:;::;
25
0
)(
0
q; a:::
....c
., .,a..u
50
L.
75
~~ \,
.~
100
10-7
10-6
Methacholine (mil) Fig. 5. Concentration-response curves to methacholine for mesenteric arteries from pregnant (n = 6, solid triangles) and nonpregnant (n = 6, solid circles) rats. Responses are expressed as a percentage of relaxation from phenylephrine preconstricted levels. Data represent mean ± SE.
Similar data from whole-animal models have further supported this role of vasodilatory prostaglandins as mediators of the blunted pressor response during pregnancy.ID However, more recent studies have resulted in
1696 Davidge and Mclaughlin
conflicting data. In chronically instrumented. trained. conscious animals treatment with prostaglandin inhibitors was not able to reverse the blunted pressor response to adrenergic vasoconstriction observed during pregnancy.4. 11 In fact. with isolated vessels there is actually very little direct evidence that vasodilatory prostaglandins are responsible for decreased sensitivity to adrenergic vasoconstriction. In aortas from guinea pigs and rats. prostaglandin inhibitors did not change the contractile response to adrenergic agents in either pregnant or nonpregnant animals. 4. 9 The results from our study with resistance-sized mesenteric arteries are in agreement with the isolated aortic ring studies. The prostaglandin products of the cyclooxygenase pathway do not appear to modulate acutely the in vitro vascular reactivity of arteries from either the pregnant or nonpregnant rats. However. our study design did not test the possibility that alterations in prostaglandin metabolism during pregnancy may have chronic effects on vascular smooth muscle function that are not apparent with acute cyclooxygenase inhibition. Another important vasorelaxant of the vessel wall is EDRF. which may have a potential role as a mediator for the blunted vasoconstrictor response observed during pregnancy. In both in vivo and in vitro systems. EDRF is known to play an important role in controlling blood vessel reactivity. It has been reported that inhibition of EDRF results in an increased sensitivity to adrenergic vasoconstriction. 19. 20 However, the role of EDRF modulation of a blunted vasoconstrictor response during pregnancy remains controversial. In conscious. instrumented nonpregnant and pregnant rats. a bolus infusion of Nw-nitro-L-arginine (an inhibitor of EDRF synthase) produced similar increases in blood pressure. 12 These data suggested that pregnancy does not result in a greater role for EDRF in the maintenance of basal blood pressure. Similar results have been reported in vitro. Parent et aI.' reported that the endothelium of the superior mesenteric artery of nonpregnant and pregnant rats similarly influenced the response to adrenergic vasoconstriction. It has also been shown that neither EDRF inhibition nor endothelium removal results in a greater adrenergic vasoconstrictor response in the aorta of the pregnant rat. 5 . 9The results from our study with resistance-sized mesenteric arteries are in agreement with the aortic ring studies in that inhibition ofEDRF or endothelial cell removal had similar increases in phenylephrine sensitivity in arteries from the pregnant and nonpregnant rats. This suggests that there is a similar endothelium-dependent modulation to a vasoconstrictor response for arteries from both the pregnant and nonpregnant rats. Contrary to the data with arteries from rats, endothelial removal or inhibition of EDRF augmented the norepinephrine
December 1992 Am J Obstet Gynecol
sensitivity more in isolated aortas and uterine arteries from pregnant guinea pigs than it did from the nonpregnant animals. 6 • 8 Because there are elevated plasma and urinary levels of cyclic guanosine 5' -monophosphate (the second messenger involved in the action ofEDRF) during gestation in the rat. 21 we had expected that the blunted response to phenylephrine during pregnancy would be caused in part by augmented EDRF activity. Although our results do not support this. it is possible that our findings are a function of the study design. If the elevated cyclic guanosine 5' -monophosphate levels indicate enhanced EDRF synthesis during pregnancy. the isolated arteries from the pregnant rats may have become L:-arginine deficient over the course of the study. Inhibition of nitrate synthase would then underestimate the contributory role of EDRF to the phenylephrine response. L-Arginine deficiency in the arteries from the pregnant rats is unlikely in our experiments. because the vessels maintained their responses to repetitive methacholine relaxation. It would be interesting to determine if pregnancy induces an isoform of nitrate synthase in the endothelium or smooth muscle that differs from the constitutive enzyme. Finally, the effect of chronic EDRF inhibition (or combined blockade of EDRF and cyclooxygenase) in vitro was not examined. Such inhibition may be required to reverse possible longer term effects of EDRF on the vascular smooth muscle. For example. nitric oxide has recently been shown to react with membrane-bound thiol groups on the N-methyl-D-aspartate (NMDA) subtype of glutamate receptor, resulting in a persistent blockade of NMDA responses!· There may be a role for EDRF in modulating vasoconstrictor responses in vivo, one that is not apparent when examined in vitro. These in vitro studies are conducted under conditions of no flow; however, flowinduced shear stress in vivo can cause production and release of ED RF from the endothelium.23 Therefore, in addition to investigating the gestational effect on basal endothelial influence, we were also interested in the gestational effect on stimulated endothelium-dependent relaxation. The effect of gestation on endothelium-dependent vasorelaxation has been investigated by looking at stimulated EDRF-mediated responses in preconstricted arterial segments or by measuring the accumulation of cyclic guanosine 5'-monophosphate in isolated arterial preparations. Again the data are conflicting as to the role EDRF has in modulating vasoactivity during pregnancy. In isolated aortas from pregnant rats cyclic guanosine 5' -monophosphate production in response to the addition of methacholine was increased compared with that in nonpregnant rats. 24 Furthermore, preconstricted uterine arteries from pregnant guinea
Volume 167 Number 6
pigs are more sensitive to endothelium-dependent relaxations, as were renal interlobar arteries from pregnant rats. 25 • 26 However, there was no difference in the endothelium-dependent relaxation of preconstricted aortas from virgin and pregnant rats when the constricting agent was norepinephrine or phenylephrine,s. 12 although there was a difference if the arteries were preconstricted with the thromboxane A2 sympathomimetic U46619. 5 In our laboratory we have observed that the potential for endothelium-dependent relaxation becomes greater with decreasing artery size, on the basis of studies with the use of the aorta and the superior and first-, secondo, and third-order branches of the mesenteric arteries (unpublished observations). Therefore the contribution that the endothelium makes in modulating vasoactivity may be more important in the resistance-sized arteries than that of the conduit arteries such as the aorta. The results from our study demonstrated that preconstricted mesenteric arteries from pregnant rats are more sensitive to relaxation by methacholine than are arteries of the nonpregnant rats. It is likely that this relaxation induced by methacholine was EDRF dependent because the relaxation was inhibited by endothelium removal or by LNAME. Furthermore, the enhanced relaxation to methacholine in arteries from the pregnant rats is most likely due to a greater release of EDRF. This can be proposed because there was no difference in relaxation response to the endothelium-independent agonist sodium nitroprusside from arteries of the pregnant and nonpregnant rats. Whether other nonmuscarinic, endothelium-dependent agents produce a similar potentiated response in these arteries from pregnant rats is unknown. In conclusion, these studies demonstrated that the decreased sensitivity to adrenergic agonists during pregnancy in the mesenteric vasculature is not modulated acutely by prostaglandin products of the cyclooxygenase pathway or by EDRF. However, the potential for endothelium-dependent relaxation is augmented in pregnancy. Because our data indicate that the vasoconstrictor response to phenylephrine is reduced by late gestation, this may reflect a generalized reduction in the smooth muscle response to vasoconstrictors. It is possible that a reduced vascular smooth muscle response to adrenergic stimulation and an increased capacity to release EDRF interact to contribute to the systemic vasodilatation observed during pregnancy. REFERENCES I. de Swiet M. Cardiovascular system. In: Hytten F, Chamberlain G, eds. Clinical physiology in obstetrics. Boston: Blackwell Scientific, 1980:3-42. 2. Gant NF, Daley GL, Chand S, Whalley P], MacDonald PC. A study of angiotensin II pressor response throughout primigravid pregnancy.] Clin Invest 1973;52:2682-9.
Endogenous modulation of vasoconstriction
1697
3. Vanhoutte PM, Rubanyi GM, Miller VM, Houston DS. Modulation of vascular smooth muscle contraction by the endothelium. Ann Rev Physiol 1986;48:307-20. 4. Harrison GL, Moore LG. Blunted vasoreactivity in pregnant guinea pigs is not restored by meclofenamate. AM] OSSTET GYNECOL 1989; 160:258-64. 5. ]ansakul C, Boura ALA, King RG. Effects of endothelial cell removal on constrictor and dilator responses of aortae of pregnant rats.] Auton Pharmacol 1989;9:93-101. 6. Moore LG. Endothelial contribution to decreased contractility in isolated aortic rings from pregnant compared to nonpregnant animals [Abstract). FASEB] 1989;3:A259. 7. Parent A, Schiffrin EL, St-Louis]. Role of the endothelium in adrenergic responses of mesenteric artery rings of pregnant rats. AM] OSSTET GYNECOL 1990;163:229-34. 8. Weiner C, Liu KZ, Thompson L, Herrig ], Chestnut D. Effect of pregnancy on endothelium and smooth muscle: their role in reduced adrenergic sensitivity. Am] Physiol 1991 ;261:H 1275-83. 9. St-Louis], Sicotte B. Prostaglandin- or endothelium-mediated vasodilation is not involved in the blunted responses of blood vessels to vasoconstrictors in pregnant rats. AM] OSSTET GYNECOL 1992; 166:684-92. 10. Pallor MS. Mechanism of decreased responsiveness to angiotensin II, NE and vasopressin in pregnant rats. Am] Physiol 1984;247:HI00-8. 11. Conrad KP, Colpoys MC. Evidence against the hypothesis that prostaglandins are the vasodepressor agents of pregnancy: serial studies in chronically instrumented, conscious rats.] Clin Invest 1986;77;236-45. 12. Umans ]G, Lindheimer MD, Barron WM. Pressor effect of endothelium-derived relaxing factor inhibition in conscious virgin and gravid rats. Am ] Physiol 1990;259: F293-6. 13. Crandall ME, Keve TM, McLaughlin MK. Characterization of norepinephrine sensitivity in the maternal splanchnic circulation during pregnancy. AM ] OBSTET GYNECOL 1990; 162: 1296-30 1. 14. McLaughlin MK, Keve TM. Pregnancy-induced changes in resistance blood vessels. AM ] OBSTET GYNECOL 1986; 155: 1296-9. 15. Osol G, Cipolla M, Knutson S. A new method for mechanically denuding the endothelium of small (50-150 J.l.m) arteries with a human hair. Blood Vessels 1989;26:320-4. 16. Hill A. The combination of hemoglobin with oxygen and carbon monoxide. Biochem] 1913;7:471-80. 17. Goodman RP, Killam AP, Brash AR, Branch RA. Prostacyclin production during pregnancy: a comparison of production during normal pregnancy and pregnancy complicated by hypertension. AM] OBSTET GYNECOL 1982;142: 817-22. 18. Everett RB, Worley R], MacDonald PC, Gant NF. Effect of prostaglandin synthetase inhibitors on pressor response to angiotensin II in human pregnancy. ] Clin Endocrinol Metab 1978;46:1007-10. 19. Martin W, Furchgott RF, Villani GM, ]othianandan D. Depression of contractile responses in rat aorta by spontaneously released endothelium-derived relaxing factor. ] Pharmacol Exp Ther 1986;237:529-38. 20. Fukuda S, Matsumoto M, Nishimura N, et al. Endothelial modulation of norepinephrine-induced constriction of rat aorta at normal and high CO 2 tensions. Am ] Physiol 1990;258:H 1049-54. 21. Conrad KP, Vernier KA. Plasma level, urinary excretion, and metabolic production of cGMP during gestation in rats. Am] Physiol 1989;257:R847-53. 22. Lei SZ, Pan Z-H, Aggarwal SK, et al. Effect of nitric oxide production on the redox modulatory site of the NMDA receptor-channel complex. Neuron 1992;8:1087-99. 23. Furchgott RF, Vanhoutte PM. Endothelium-derived relaxing and contracting factors. FASEB] 1989;3:2007-18.
Davidge and Mclaughlin
24. Whittemore SL, Conrad KP. Stimulation of cyclic guanosine 5' -monophosphate (cGMP) production by methylcholine differs in aortae from pregnant and virgin rats [Abstract]. Physiologist 1989;32: 186. 25. Weiner CP, Martinez E, Zhu LK, Ghodsi A, Chestnut DH. In-vitro release of endothelium-derived relaxing factor by acetylcholine is increased during the guinea pig pregnancy. AM] OSSTET GYNECOL 1989;161:1599-605.
December 1992 Am J Obstet Gynecol
26. Griggs KC, McLaughlin MK. Endothelial modulation of vasoconstrictor response of isolated renal interlobar arteries from pregnant and nonpregnant rats [Abstract]. In: Proceedings of the thirty-eighth annual meeting of the Society for Gynecologic Investigation, San Antonio, Texas, March 20-23, 1991. San Antonio: Society for Gynecologic Investigation, 1991.
A unique hypertonic response to hypotonic infusion in the pregnant ewe David R. Powers, MD,· and Robert A. Brace, PhDb
San Diego, California OBJECTIVE: The purpose of this study was to compare the responses of the maternal ewe to intravenous volume expansion with either sufficient lactated Ringer's solution to elevate maternal venous pressure or sufficient hypotonic fluid to reduce blood osmolality. STUDY DESIGN: Chronically catheterized pregnant sheep were intravenously infused over 4 hours with either commercial lactated Ringer's solution (5.55 ± 0.50 Uhr, 255 mOsm/kg, mildly hypotonic) or diluted Ringer's solution (2.04 ± 0.27 Uhr, 150 mOsm/kg, markedly hypotonic). Data were statistically analyzed with two- and three-factor analyses of variance and bivariate regression analysis. RESULTS: During the mildly hypotonic infusion (n = 8) the maternal blood osmolality changes were - 5.1 ± 1.2, + 2.7 ± 1.0 and + 6.8 ± 1.1 mOsm/kg at 1 and 4 hours of infusion and 1 hour after the infusion. In four of the eight animals in this group profuse diarrhea developed. During the markedly hypotonic infusion (n = 11) the maternal blood osmolality changes were - 9.9 ± 1.1, -15.9 ± 2.5, and -10.4 ± 2.2 mOsm/kg at 1 and 4 hours of infusion and 1 hour after the infUSion. Although urine osmolalities were significantly less than the osmolality of the infusate in both groups, only during the mildly hypotonic infusion was there a net loss of free water by the kidneys. The renal free water loss, the venous pressure increase, and the blood osmolality decrease were not significantly different whether diarrhea did or did not develop. CONCLUSION: The infusion of large volumes of mildly hypotonic Ringer's solution to the pregnant ewe produces a paradoxic increase in maternal plasma osmolality as a result of the excretion of large volumes of free water by the kidneys, and if the venous pressure is increased more than about 6 mm Hg with this infusion, diarrhea develops in the animals. (AM J OesTET GVNECOL 1992;167:1698-709.)
Key words: Pregnancy, osmolality, ewe, free water clearance, diarrhea Maternal volume status changes dramatically during human pregnancy. Blood volume increases by about
From the Division of Perinatal Medicine, Department of Reproductive Medicine, University of California at San Diego, a, b and the Division of Nephrology, Department of Internal Medicine, University of California at Irvine. a Supported in part by National Institutes of Health grants HD20295 and HD23724 from the National Institute of Child Health and Human Development. Presented at the Thirty-eighth Annual Meeting of the Society for Gynecologic Investigation, San Antonio, Texas, March 21-23,1991. Reprint requests: David R. Powers, MD, Department of Reproductive Medicine, University of California at San Diego, La Jolla, CA 92093-0802.
6/6/41716 1698
1.5 L, and the change in plasma volume accounts for the majority of the increase. I - 3 Furthermore, a failure of maternal plasma volume to expand has been associated with disease states such as preeclampsia,., 5 chronic hypertension,5. 6 intrauterine growth retardation, 7 and 0ligohydramnios. 8 If maternal volume status is indeed compromised in these conditions, the result could be a decrease in maternal hydrostatic pressure perfusing the placenta or an increase in osmotic pressure of maternal plasma, both potentially decreasing fluid movement to the fetus. In spite of recent studies suggesting that maternal volume expansion may be of benefit in the treatment of preeclampsia9 or oligohydramnios,1O little is currently known about the effects of fluid therapy on