Pregnancy-induced ANDREW
D. HULL,
changes in ovine cerebral arteries DON
M. LONG,
LAWRENCE
D. LONGO,
AND
WILLIAM
J. PEARCE
Division of Perinatal Biology, Departments of Physiology and Obstetrics and Gynecology, Loma Linda University School of Medicine, Loma Linda, California 92350 Hull, Andrew D., Don M. Long, Lawrence D. Longo, and William J. Pearce. Pregnancy-induced changes in ovine cerebral arteries. Am. J. Physiol. 262 (Regulatory Integrative Comp. Physiol. 31): R137-R143, 1992.-We examined the effects of pregnancy on the ovine cerebral vasculature by comparing several characteristics of isolated endothelium-intact segments of three intracranial arteries including the middle cerebral (MCA), posterior communicating (PC), and basilar (BAS) arteries taken from pregnant sheep (138143 days gestation, term ~145 days) and nonpregnant controls. For comparison, segments of the extracranial common carotid (COM) artery were also studied. With pregnancy, vessel water content increased (5.458%) in all arteries except the PC. Additionally, cellular protein content increased in all arteries (4.4~50.0%). Arterial stiffness, as determined by passive stress-strain determinations, was significantly decreased during pregnancy in the MCA but not in the larger arteries. Maximum contractile responses, when normalized to vessel wall cross-sectional area, were consistently greater in arteries from pregnant than in those from nonpregnant animals (lo&49.7%). Relaxation to the endothelium-independent guanylate cyclase stimulator Snitroso-N-acetyl penicillamine (SNAP) increased with pregnancy only in the distal MCA (=17%). Endothelium-dependent relaxation to the calcium ionophore A23187 decreased only in the larger and more proximal COM (-39%). Thus pregnancy was associated with an increase in production of contractile force, a decrease in peripheral vascular stiffness, a decrease in the relaxant response to A23187 in the COM, and an increase in the relaxant response to SNAP in the MCA. Together, these findings indicate that pregnancy has widespread and important vessel specific cerebrovascular consequences that affect not only arterial composition, but also contractility and endothelial reactivity. cerebrovascular circulation; endothelium; vascular smooth muscle PREGNANCY induces major changes in the maternal cardiovascular system including an increase in blood volume of up to 45%, a doubling of cardiac output, a fall in
peripheral resistance, and major alterations in regional blood flow (11). Individual vascular beds, however, vary considerably in response to pregnancy (1, 23). Understandably, much attention has focused on the effects of pregnancy on the uteroplacental and renal circulations. However, examination of other vascular beds is important in determining
the more widespread
adaptations
of
the circulatory system to gestation. Elucidation of these changes should enhance our understanding of both normal and complicated pregnancy. One area worthy of examination during pregnancy is endothelial function. The endocrine and paracrine influences of the vascular endothelium are now widely recognized (27). The possibility that endothelial dysfunction is important in the pathogenesis of many disorders of pregnancy, including pre-eclampsia (21), makes the study of normal endothelial function and response during pregnancy of paramount importance.
In the present studies, we have examined adaptation of the cerebral circulation to pregnancy. Although the cerebral circulation has been held to be little affected by pregnancy (16, 23), relatively few studies of the effects of pregnancy on this vascular bed have been conducted. Given the possible relations between vessel composition, structure, and function, we have examined water and protein content, thickness and stiffness, and reactivity to both endothelium-dependent and endothelium-independent vasodilators. Recognizing that any alterations in function related to pregnancy may be vessel specific, we have examined arteries from four levels of the cerebrovascular tree. METHODS From pregnant sheep (n = 12) at 138-143 days gestation (term ~145 days) and from nonpregnant control animals (n = 48) we obtained complete segments of three intracranial arteries including basilar (BAS), posterior communicating (PC), and middle cerebral arteries (MCA). For comparison, segments of the extracranial common carotid (COM) artery were also studied. The nonpregnant animals were nulliparous young adults, confirmed by the United States Department of Agriculture to be disease free, obtained from a local abbatoir. After initial dissection and removal of excess adipose and connective tissue, we cut the vessels into individual rings 3 mm long (COM and PC) and 5 mm long (BAS and MCA). We determined vessel wall thicknesses using projection microscopy in fresh unfixed representative transverse sections of each artery as previously described (19). The anatomical integrity of the vascular endothelium was established in samples of arteries from representative animals using scanning electron microscopy, and in all cases the endothelium was found to be physically intact. In separate parallel functional experiments, responses to a submaximal dose (1 PM) of A23187, an endothelium-dependent vasodilator, were consistent and reproducible. Together, these data indicate that the endothelium was intact and functional in the arteries used in the present studies. Water and protein determinations. We determined relative vessel water and protein content as previously described (20). In brief, we weighed blotted arterial segments before and after 48 h desiccation at 50°C to determine water content and assayed vessel protein content using an extraction designed to exclude connective tissue and structural proteins (8, 25). We quantified protein using the Bradford Coomassie Brilliant Blue assay (5). Contractility experiments. We mounted each vessel segment on paired wires between a low compliance force transducer (Kulite BG-lo), and a post attached to a micrometer used to vary baseline tension. We equilibrated the arteries at 38.5OC for 30 min in a bicarbonate Krebs solution containing (in mM) 122 NaCl, 25.6 NaHCOs, 5.56 dextrose, 5.17 KCl, 2.49 MgSOI, 1.60 CaC12, 0.114 ascorbic acid, and 0.027 disodium EDTA, continuously bubbled with 95% Og-5% COZ. We obtained micrometer readings, and thereby diameter measurements, for each arterial segment under unstressed conditions (CO.1 g tension), and similarly at an optimum resting tension of 0.5 g
0363-6119/92 $2.00 Copyright 0 1992 the American Physiological Society
R137
Downloaded from www.physiology.org/journal/ajpregu by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 14, 2019.
R138
PREGNANCY
AND OVINE
for the cerebral vesselsand 1 g for the carotid arteries. Vessel diameterwascalculatedastwice the measureddistancebetween mounting wires divided by 7~ We calculated force per unit cross-sectionalarea,asdescribedpreviously (19), asthe product of maximum tension and the accelerationdue to gravity divided by cross-sectionalarea. Cross-sectionalarea is calculated as twice the product of vessellength and vesselwall thickness at optimal tension, which is derived by dividing the measured vessel diameter at rest by the measuredvessel diameter at optimum tension and multiplying this value by the vesselwall thickness measuredat rest. We first contracted the arteries by exposing them to an isotonic potassiumKrebs solution containing 122 mM K+ and 31 mM Na’ (the composition of this solution was identical to that of our “normal” sodium Krebs solution except that there was an equiosmolarexchangeof potassiumfor sodium). After peak tensionswere reached,we washedthe vesselswith normal sodiumKrebs solution and allowed them to return to baseline levels of tension for 30 min. We then induced a secondcontraction using a mixture of 10 PM serotonin and 20 PM histamine. Previous studies have shown this mixture to produce maximal contractions with greater stability than any other method of contraction in the ovine cerebral arteries (18). During all contractility experiments, we continuously digitized, normalized, and recordedcontractile tensions using an on-line computer. Determination of stress-strainrelations. We mounted vessel segmentsasdescribedabove, and obtained micrometer readings of vesseldiameter at near zero tension and at baselinetensions of 0.5 g (cerebral arteries) and 1 g (COM). These micrometer readingsenabledus to calculate true unstressedresting vessel diameter (D,) and vesseldiameter at baselinetension (&). We then carried out a seriesof potassiumdepolarization contractions at resting vesseldiameters varied over the range D/Db 0.5-1.7, and recorded the resultant maximum active tensions. From thesedata, we wereable to plot active stress/strain curves for each vessel,using true values of wall stress,thus allowing the determination of optimum resting diameter (Do& for each vessel.This measurementof DOptis conceptually equivalent to the measurementof optimal length, Lo, usedby other authors. Once DOpthad been determined, we froze the vesselswhile still mounted on their wires by immersion in liquid nitrogen to eliminate any active component of wall stress.After thawing, we generated a passive stress-strain curve, again by varying vesseldiameter over the rangeD/DOpt0.5-1.7 and recording the resultant passivetension at each diameter. Vasorelaxantstudies.We studied the relaxant responsesto two agents, A23187 and S-nitroso-N-acetyl penicillamine (SNAP). The calcium ionophore A23187 is a receptor-independent stimulator of endothelium-derived relaxant factor (EDRF) releasefrom the vascular endothelium (30). Its use therefore served as an indicator of the endothelium-evoked vasorelaxation through stimulation of guanylate cyclase and accumulation of guanosine 3’,5’-cyclic monophosphate (cGMP). In contrast, the nitrosothiol SNAP acts directly on the vascular smooth muscle independent of the endothelium. The rapid releaseof nitric oxide from SNAP producesmaximal stimulation of guanylate cyclase,maximal production of cGMP, and hencemaximal relaxation (12). We contracted the vesselswith serotonin and histamine as describedabove and, after establishment of stable tone, added either 1 PM A23187or 10 PM SNAP and recorded the resulting changesin tension producedby eachvessel.We determined the relaxant responsefor each vesselby integrating the area under each tension-time course curve. We expressedthe average of these values as a percentage of maximum tension, to give averagepercent relaxations for each vesseltype to each agent. All methodsand nroceduresusedin this studv were annroved
CEREBRAL
ARTERIES
by the animal care committee at our institution. Statistics. Throughout the text, values are reported asmeans t SE, unlessotherwise indicated. Where a single animal contributed more than one vesselsegmentto any particular protocol, the individual results were averaged to a single value before analysis of group statistics. A total of 416 nonpregnant and 174 pregnant artery segmentswere used.For all statistical comparisons either a l-way or a 2-way analysis of variance (ANOVA) was employed. Integrals of the tension-time curves were determined by Simpson’sapproximation. Comparisonsof ratios were made using a Behrens-Fischer analysis. All probabilities ~0.05 were regardedas significant. RESULTS
Vessel cellular protein and water content. In both the nonpregnant and pregnant animals, there was no significant difference in protein content between arteries (Fig.
1). Only in the COM was pregnancy significant
associated with -a
increase in protein content. Water content
was not significantly different between arteries in either group (Fig. 1). With pregnancy there was a significant increase of ~5.5% in all but the PC. Vessel wall thickness. With pregnancy, there was a decrease of 6.8 t 0.9% in the COM and slight increases in wall thickness in all the intracranial cerebral arteries in the pregnant animals (Table 1). However, none of these changes were statistically significant. Contractility. The maximal contractile responses to 122 mM potassium were significantly higher in the COM than in the smaller intracranial arteries in both nonpregnant (15.2 t 0.8 g) (P < 0.0001 ANOVA) and pregnant (18.5 t 1.8 g) animals (P < 0.0001 ANOVA). Between-artery differences were not significant in either group for any other artery (nonpregnant: BAS 3.0 t 0.2 g, PC 2.6 t 0.2 g, MCA 2.9 t 0.2 g; pregnant: BAS 2.9 t 0.3 g, PC 3.0 t 0.4 g, MCA 3.6 t 0.5 g). Pregnancy was H
Nonpregnant
N
Pregnant
*
COM
_ BAS
-
PC
- MCA
-I
Fig. 1. Pregnancy-induced changes in cellular protein and water content in ovine cerebral arteries. Protein contents were calculated as a percentage of vessel dry weight, and water contents were calculated as the percent differences between wet and dry weights. Both are shown here as means k SE for the numbers of animals indicated at the base of each bar. COM, common carotid artery; ‘BAS, basilar artery; PC, posterior communicating artery; MCA, middle cerebral artery, ‘P < 0.05 comnared with nonnreenant.
Downloaded from www.physiology.org/journal/ajpregu by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 14, 2019.
PREGNANCY
Table
AND OVINE
1. Vessel wall thickness Artery Common carotid
Posterior communicating
Basilar
Middle cerebral
Nonpregnant 935k17 (48) 112t2 (44) 139t5 (44) 123k2 (44) Pregnant 871_t39 (12) 113*5 (11) 144,tlO (12) 13027 (12) Vessel wall thicknesses are given above in microns and are expressed as means t SE for the numbers of animals given in parentheses. Changes with pregnancy were not significant for any artery type.
(Yn 2E Ow8 *c 0.6 3 04 5 0.2 0
0.6
E = 0.4 E 5 0.2 00
Carotid
Basilar
l
8Q) O 8 .-s! 5 0.8
a
Common
Pouterior
Communicating
Middle
Cerebral
0.8 0.6 0.4 0.2 0
Fig. 2. Pregnancy-induced changes in maximal cerebrovascular stress production. For each artery type, the maximum force produced by exposure to either 122 mM potassium, or 10 PM serotonin with 20 PM histamine, was normalized relative to vessel wall cross-sectional area. The resultant values of maximum active stress are given here as means and standard errors for the numbers of animals indicated at the base of each bar. §P < 0.05 compared with all other nonpregnant vessels; tP < 0.05 compared with all other pregnant vessels; *P < 0.05 compared with nonpregnant.
associated with a significant rise only in the COM (21.8 t 0.9%) (P < 0.05, ANOVA). Figure 2 shows maximum contractile force normalized relative to vessel wall cross-sectional area. In nonpregnant animals, maximal wall stress was lowest in the COM (0.31 t 0.01 x lo6 dyn/cm2) and increased in the more peripheral vessels: BAS (0.36 t 0.03 x lo6 dyn/ cm2), MCA (0.45 t 0.03 x lo6 dyn/cm2), and PC (0.55 t 0.04 x lo6 dyn/cm2). Wall stress was significantly higher in the PC compared with each of the other arteries. In the pregnant group, the PC again produced significantly more wall stress (0.80 $- 0.1 X lo6 dyn/cm’) than the other vessels, COM (0.43 t 0.02 x lo6 dyn/cm2), BAS (0.54 t 0.07 x lo6 dyn/cm2), and MCA (0.49 t 0.07 x lo6 dyn/cm2). Increases in maximal wall stress associated with pregnancy were significant in the COM (41.2 t 0.3%), BAS (49.7 t 0.6%), and PC (46.7 t 0.6%) but were not in the MCA (10.1 t 2.4%). When the maximal contractile response to a mixture of 10 PM serotonin and 20 PM histamine was expressed as a percentage of that induced by depolarization with 122 mM potassium, a similar pattern across the vascular tree was seen in vessels from both nonpregnant and pregnant animals. The greatest sensitivity to the agonists used was seen in the PC (nonpregnant 100.4 t 5.4%, pregnant 101.5 $- 12.8%) and the least in the COM
CEREBRAL
R139
ARTERIES
(nonpregnant 30.5 t 2.3%, pregnant 25 t 2.7%). The MCA (nonpregnant 68.7 t 5.6%, pregnant 69.6 t 7.2%) and BAS (nonpregnant 54.3 t 4.4%, pregnant 70.4 k 7.3%) showed intermediate values. Between-vessel differences were significant in all cases except in the pregnant BAS vs. MCA. None of the changes seen with pregnancy were statistically significant. Vessel diameters and passive wall stiffness. Not surprisingly, in both the nonpregnant and pregnant groups, resting vessel diameters were lower in the more periphera1 vessels (Fig. 3). None of the changes in resting diameter or optimum diameter seen with pregnancy were statistically significant (ANOVA). In the nonpregnant animals there were no significant differences in passive compliance between arteries (Fig. 4). However, in the pregnant group the MCA was significantly less stiff than the other vessels, which all had similar compliance values. Pregnancy produced a significant change in compliance only in the MCA. Vasorelaxant studies. The relaxant response to A23187 was similar in all four artery types obtained from the nonpregnant animals, ranging from 26.3 t 5.5% (MCA) to 37.6 t 6.5% (PC) (Fig. 5). Only the pregnancy-associated reduction in relaxation response to A23187 in the COM (from 33.2 t 3.5% to 20.1 t 2.5%) was significant. The degree of vasorelaxation produced by SNAP was significantly greater than that produced by A23187 in all vessels in both pregnant and nonpregnant groups. The smaller, more peripheral arteries relaxed significantly more (~75%) than the proximal COM (~40%). Only in the MCA was a significant effect of pregnancy seen on the ability of SNAP to relax the vessels (an increase from 64.9 t 2.2 to 75.9 t 2.4%). An estimate of the relative ability of the endothelium to recruit vasorelaxation was obtained by calculating the ratio of relaxation produced by A23187 to that produced n E E 5 5 E m E
Common
Carotid
Basilar
a
2
6
1.5
1,
’ 2 0 _
0.5
0
Posterior
Do
Communicating
Dopt.
Middle
Do
Cerebral
Dopt.
Nonpregnant
Pregnant RI Fig. 3. Pregnancy-induced changes in resting and optimum diameters in ovine cerebral vessels. Resting diameter at zero tension (Do) and optimum diameter (Do&, determined from active length-tension curves as described in the text, are given here for each vessel type. Values are means ,t SE for the number of animals given at the base of each bar. No significant differences between pregnant and nonpregnant animals were observed.
Downloaded from www.physiology.org/journal/ajpregu by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 14, 2019.
R140
PREGNANCY +
Common
Nonpregnant
6
AND OVINE
ARTERIES
s
Pregnant
Carotid
CEREBRAL
n
s 2
Basilar
s 5
N
Nonpregnant
Pregnant
60 1
a
v)
0.4
0.6
0.8
1
1.2
0.4
0.6
0.8
1
1.2
v) d) CI
Posterior
Communicating
Middle
Cerebral
1
n
.p 0.8 s E 0.6 0.4
0.6
0.8
1
1.2
0.4
0.6
0.8
1
1.2
L/Lo Fig. 4. Pregnancy-induced changes in passive stress-strain relations in ovine cerebral arteries. For each artery type, passive stress-strain curves are given for the range L/L,, 0.4-1.2, where L is true vessel diameter and L, is optimal vessel diameter. Values are given as means; errors bars are shown where larger than the symbols used. *P < 0.05 compared with all other nonpregnant and pregnant vessels; n = 6 for all but the carotid arteries from pregnant animals, where n = 4.
by SNAP (Fig. 5). There was a significant decrease in the ratio in the COM from 0.78 t 0.14 in the nonpregnant animals to 0.50 t 0.12 in the pregnant group. No consistent effect of pregnancy was seen in the smaller arteries. DISCUSSION
In humans, pregnancy has been shown to have little effect on global cerebral blood flow (16). This absence of effect has also been demonstrated in sheep by Rosenfeld (23) using radionuclide-labeled microsphere techniques. Overall, ovine cardiovascular adaptations to pregnancy are generally less pronounced than those seen in humans (26). Nonetheless, it is apparent from the present results that the ovine cerebral vasculature is significantly affected by pregnancy. It follows, therefore, that the vascular changes here described must be in balance with other circulatory changes associated with pregnancy to maintain constancy of cerebral perfusion. Total body water, in particular the extracellular fluid volume, is well known to increase in pregnancy. This increased state of hydration is shown in all but the PC (Fig. 1) with increases of =5.5% in vessel water content. Although their data were not statistically significant, Griendling et al. (10) found increases in water content of 2% in carotid and 7% in uterine arteries taken from pregnant sheep compared with nonpregnants. Because relative water content increases during pregnancy, it
a 3
0.4
Fi 2
0.2 0
COM
BAS
.
PC
. MCA
’
Fig. 5. Pregnancy-induced changes in relaxation response to A23187 and SNAP. For each artery type, percent relaxation produced by exposing previously contracted vessels to either 1 PM A23187 or 10 /IM SNAP is shown. Also given is the ratio of A23187-induced relaxations to SNAP-induced relaxations. The relaxation responses are shown as means k SE for the number of animals indicated at the base of each bar. COM, common carotid artery; BAS, basilar artery; PC, posterior communicating artery; MCA, middle cerebral artery. *P < 0.05 compared with nonpregnant.
follows that other vessel wall constituents may decrease. Although we have not specifically addressed this possibility, others (10) have found no significant change in ovine carotid artery fat or connective tissue content with pregnancy. Another possibility is that the mucopolysaccharide content is altered, as suggested by others in other vascular beds (22). Estrogen, derived from dehydroepiandrosterone sulfate produced by the fetal adrenal, stimulates the production of renin substrate and stimulates aldosterone secretion, thus encouraging sodium and water retention. Progesterone similarly increases renin activity and aldosterone levels. Thus the driving force for the changes in vessel hydration is likely to be hormonal. The lack of change in water content in the PC may reflect differences in numbers or level of activation of sex hormone receptors at different sites in the vasculature. The fraction of protein we assayed included only cellular and enzymatic proteins, and thus the data given in Fig. 1 do not reflect any pregnancy-related alterations in vascular connective tissue content. Minor increases in relative protein content were seen in the smaller arteries, but a significant rise of 50% was seen in the COM. As with our measurements of water content, the protein measurements were relative to total dry weight, implying
Downloaded from www.physiology.org/journal/ajpregu by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 14, 2019.
PREGNANCY
AND
OVINE
that some other cellular constituent is decreasing its relative contribution to dry mass. Annibale et al. (3), using a different assay technique, found an increase in vessel protein content with pregnancy in ovine uterine but not carotid or renal arteries. The reason for the difference between this and our findings may be that their values were expressed as percentages of wet weight, and the changes in vessel water content with pregnancy may vary in a site-specific manner. No significant alterations in vessel wall thickness were associated with pregnancy, nor were any clear patterns of variation discernable in the arteries studied. No attempt was made to quantify the relative contribution each component of the vessel wall made to overall thickness or changes therein. A reduction in both media and overall wall thickness in 165-pm mesenteric arteries (approximately half the diameter of our smallest artery, the MCA) was found in pregnant rats at term by other investigators (17). It may be that such structural alterations are limited more to resistance-sized vessels remote from the uteroplacental circulation. In contrast, the dramatic growth in the uterine arteries with pregnancy is well recognized. Several investigators have demonstrated in a variety of vessels in several different species that optimal resting tension increases with pregnancy (17, 28, 29). Figure 3 indicates that optimal vessel diameter, and by extension optimal resting tension, does not change uniformly across the vascular tree with pregnancy. However, in this study, none of the changes seen were statistically significant. The first published study of changes in arterial mechanics with pregnancy showed no effect on the passive compliance of rabbit aorta (13). Subsequent studies of ovine carotid artery have corroborated this finding (2, 10). In contrast, studies of resistance-sized vessels have shown significant pregnancy-associated decreases in passive stiffness (17). Our findings in the COM, BAS, and PC agree with the earlier studies conducted in the larger arteries; that is, we observed no significant change in passive compliance with pregnancy (Fig. 4). In contrast, and consistent with studies in smaller arteries, we observed a significant loss of stiffness in the MCA, which in our studies was x400 pm in diameter. Aside from these apparent modest changes in stiffness, the resting tone of the arteries may have also increased, as suggested by the observed values of D, (Fig. 3). These values, which also exhibited vessel-specific changes with pregnancy, decreased in the COM, BAS, and PC but remained unchanged in the MCA. Overall, there is good evidence that the general distensibility of the circulatory system increases in pregnancy (4). What then are the underlying changes in the arterial wall that can explain the observed changes in passive mechanical characteristics? Simple histological examinations, unfortunately, do not give the answer. Danforth (6) demonstrated fragmentation of the reticulum and attenuation of the elastica in rabbit aorta, but these changes appeared to have no effect on mechanical properties (13). Studies comparing estrogen-supplemented and untreated ovariectomized animals (7) suggested that estrogen stimulates collagen and elastin degradation at
CEREBRAL
ARTERIES
R141
differential rates in rat aorta, leading to a greater proportion of elastic fibers and hence greater distensibility. However, no absolute change in collagen, elastin, or collagen/elastin ratio was found with pregnancy in ovine carotid artery by Griendling et al. (lo), although a decrease in collagen was found in the uterine artery. The present data suggest that the changes taking place in the connective tissue jackets of arteries, and in resistance vessels in particular, are more subtle than thus far suggested. Alterations in the spatial relationship of the noncontractile elements of the vessel wall in a looser, more watery ground substance may occur. Altered collagen synthesis and turnover or modification of the crosslinkages between fibers may take place. These areas clearly require further investigation. Consistent with the observed increases in cellular protein, pregnancy was also associated with an increase in contractile capacity in all vessels, as determined by potassium depolarization. When these contractile tensions were normalized for vessel wall cross-sectional area (Fig. 2), significant increases were still seen in all but the MCA. Such changes may represent alterations in the coupling of membrane depolarization and Ca2+ entry into the vascular smooth muscle cells, and the subsequent activation of the contractile process. Clearly these complex pathways are potentially subject to modification by the hormonal influences of pregnancy. Comparisons between our findings and those of others are complicated by the fact that in most studies, contractile tensions are rarely normalized for wall cross-sectional area. McLaughlin and Keve (17) found a pregnancy-related decrease in contractile force in the rat mesenteric artery, a vessel considerably smaller than any of the arteries we studied. Consistent with the findings of McLaughlin and Keve, we observed a falloff in the size of the pregnancyinduced increase in force generation in the more peripheral vessels. Although Annibale et al. (2) demonstrated an increase in wall stress with pregnancy in the ovine uterine artery, their increase of ~40% in wall stress in the COM (cf. our finding of an increase of 41.2 t 0.3%) was not statistically significant. A later study (3) showed no increase in cellularity in the pregnant ovine uterine artery to account for the increase in stress production. There was, however, a demonstrable increase in both actin and myosin content as part of a general increase in total protein and a rise in the level of myosin light chain phosphorylation after contraction with phenylephrine. The ovine renal artery did not show these findings. Relaxation responses to A23187 were similar (~30%) in the smaller arteries in both pregnant and nonpregnant animals (Fig. 5). In the COM, however, there was a significant reduction in relaxation with pregnancy. In a recent study of acetylcholine-induced relaxation in guinea pig carotid and uterine arteries (29), pregnancy was associated with a significant enhancement of endothelium-mediated vasorelaxation, an effect attributed to an augmentation of EDRF activity. Support for this interpretation came from an earlier study of castrated estrogen-treated rabbits whose femoral artery relaxation response to acetylcholine was enhanced relative to that of untreated nonpregnants (9). However, additional experiments using A23187 in the same study showed no
Downloaded from www.physiology.org/journal/ajpregu by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 14, 2019.
R142
PREGNANCY
AND OVINE
enhancement of response by estrogen treatment. It seems likely, therefore, that the greater relaxation to acetylcholine seen in both groups of experiments is an estrogenmediated receptor effect rather than a fundamental alteration in EDRF production or release. It remains possible, however, that acetylcholine and A23187 may not release the same EDRF (24). In the present data, the equivalency of the responses to A23187 in the smaller arteries of both groups suggests that EDRF production in these arteries is unaltered by pregnancy. In contrast with this finding, pregnancy was associated with a decrease in A23187-induced relaxation in the COM. The reason and mechanism for this depression remain unclear. Basal EDRF production may be increased in pregnancy, thereby reducing the apparent “relaxing capacity” of the endothelium in response to agonists. This is supported by the finding of an increase in basal cGMP levels in the pregnant ovine uterine artery (15). Equally, changes in any of a vast number of steps could explain this finding, including a decrease in the calcium sensitivity of endothelial cells, a decrease capacity for EDRF synthesis, or a limitation of EDRF precursor, to name only a few possibilities. One mechanism that cannot explain the decreased sensitivity to A23187 in the pregnant COM is that the vascular guanylate cyclase, which mediates the effects of EDRF, was somehow less sensitive to EDRF in the pregnant animals. SNAP, which releases nitric oxide and thus produces relaxation via the same pathway as that used by the endothelium (l4), produced rapid and maximal relaxation in all arteries (Fig. 5). Responses to SNAP were unaffected by pregnancy, except in the most distal artery studied, the MCA, where an enhancement was observed. Thus a decrease in sensitivity to EDRF cannot explain the decreased responsiveness to A23187 observed in the pregnant COM. Instead, a direct effect of pregnancy on endothelial synthesis and/or release of EDRF is suggested. In that endothelial dysfunction has been suggested as a possible factor in pre-eclampsia (21)) this latter suggestion warrants additional study. Conclusions. The present study demonstrates that pregnancy affects the cerebral vascular bed, despite its remoteness from the conceptus. Although increases in water, protein content, and force-generating ability were seen as a general phenomenon, other parameters changed in a highly vessel-specific manner. The passive mechanical characteristics of the larger arteries were unchanged, whereas the distal MCA showed a significant loss of stiffness. The capacity of the endothelium to elicit vasorelaxation was altered with pregnancy only in the COM, where a decrease was demonstrated. The maximal relaxation potential of the vascular smooth muscle changed only in the MCA, which showed an increase. Alterations in both the passive and active characteristics of the MCA may parallel the general decrease in peripheral vascular resistance seen elsewhere with pregnancy. Endothelial function is altered only in the COM, although the reason for this is uncertain. Clearly, the results of the present study, when combined with the work of others, indicate that pregnancy has profound effects on the vascular system. Because many of these effects are vessel and site specific, findings in one bed
CEREBRAL
ARTERIES
may not be directly extrapolated
to another.
The authors extend their appreciation to Charles White and Al van Varik for their excellent technical assistance with these studies. The authors also thank Dr. Louis Ignarro for the SNAP used in these studies. This project was supported by National Institutes of Health Grants HL-41347 and HD-03807 and Grant 90-121 from the California Affiliate of the American Heart Association. Address for reprint requests: A. D. Hull, Division of Perinatal Biology, Loma Linda University School of Medicine, Loma Linda, CA 92350. Received 25 February 1991; accepted in final form 9 August 1991. REFERENCES 1. Abramson, D. I., K. Flachs, and S. M. Fierst. Peripheral blood flow during gestation. Am. J. Obstet. Gynecol. 45: 666-671, 1943. 2. Annibale, D. J., C. R. Rosenfeld, and K. K. Kamm. Alterations in vascular smooth muscle contractility during ovine pregnancy. Am. J. Physiol. 256 (Heart Circ. Physiol. 25): H1282-H1288, 1989. 3. Annibale, Kamm.
D.
J.,
C. R.
Rosenfeld,
J.
T.
Stull,
and
K.
K.
Protein content and myosin light chain phosphorylation in uterine arteries during pregnancy. Am. J. Physiol. 259 (Cell
Physiol. 4. Bonnar,
28): C484-C489, 1990. J., I. MacGillivray, and E. M. Symonds. reactivity in pregnancy. In: Pregnancy Hypertension.
Vascular Baltimore,
MD: University Park Press, 1980. M. M. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72: 248-254, 1976.
5. Bradford,
6. Danforth,
D. N., P. Manola-Estrella,
and J. C. Buckingham.
The effect of pregnancy and of Enovid on the rabbit vasculature. Am. J. Obstet. Gynecol. 88: 952-962, 1964. 7. Fischer, G. M. In vivo effects of estradiol on collagen and elastin dynamics in rat aorta. Endocrinology 91: 1227-1232,1972. 8. Furoto, D. K., and E. J. Miller. Isolation and characterization of collagens and procollagens. Methods Enzymol. 144: 41-61,1987. 9. G&lard, V., V. M. Miller, and P. M. Vanhoutte. Effect of 17
beta-estradiol 10.
J. Pharmacol. Griendling,
on endothelium-dependent
responses in the rabbit.
Exp. Ther. 244: 19-22, 1988. K. K., E. 0. Fuller, and R. H. Cox.
Pregnancyand carotid arteries. Am. J. 17): H658-H665, 1985. Adaptations of metabolism
induced changes in sheep uterine Physiol.
248 (Heart Circ. Physiol. and G. Chamberlain.
11. Hytten, F., and nutrient Boston, MA: 12. Ignarro, L. A. L. Hyman,
requirements. In: Clinical Physiology Blackwell, 1980, p. 345-350. J.,
H. Lippton, P. J. Kadowitz,
in Obstetrics.
J. C. Edwards, W. H. Baricos, and C. A. Gruetter. Mechanism
of vascular smooth muscle relaxation by organic nitrates, nitrites, nitroprusside and nitric oxide: evidence for the involvement of Snitrosothiols as active intermediates. J. Pharmacol. Exp. Ther. 218: 739-749,198l. 13. Johnson, W. L., J. T. Conrad,
D. Whitney,
and N. Graybeal.
The effect of pregnancy and norethynodrel with mestranol (EnOvid) on length-tension relationship in the rabbit aorta. Am. J. Obstet. Gynecol. 93: 179-180, 1965. 14. Lincoln, T. M. Cyclic GMP and Pharmacol. Ther. 41: 479-502,1989. 15. Magness, R. R., C. R. Rosenfeld,
mechanisms of vasodilation.
and P. W. Shaul. Role of endothelium in the control of CAMP and cGMP production by uterine artery vascular smooth muscle in ovine pregnancy (Abstract). Proc. Sot. Gynecol. Invest. 265, 1991. 16. McCall, M. L. Cerebral blood flow and metabolism in toxaemias of pregnancy. Surg. Gynecol. Obstet. 89: 715-721, 1949. 17. McLaughlin, M. K., and T. M. Keve. Pregnancy-induced changes in resistance blood vessels. Am. J. Obstet. Gynecol. 155: 1296-1299,1986. 18. Pearce, W. J., and
S. Ashwal. Developmental changes in thickness, contractility, and hypoxic sensitivity of newborn lamb cerebral arteries. Pediatr. Res. 22: 192-196, 1987. 19. Pearce, W. J., S. Ashwal, and J. Cuevas. Direct effects of graded hypoxia on intact and denuded rabbit cranial arteries. Am.
Downloaded from www.physiology.org/journal/ajpregu by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 14, 2019.
PREGNANCY
AND OVINE
J. Physiol. 257 (Heart Circ. Physiol. 26): H824-H833, 1989. Pearce, W. J., A. D. Hull, D. M. Long, and L. D. Longo. Developmental changes in ovine cerebral artery composition and reactivity. Am. J. Physiol. 261 (Regulatory Integrative Comp. Physiol. 30): R458-R465, 1991. 21. Roberts, J. M., R. N. Taylor, T. J. Musci, G. M. Rodgers, C. A. Hubel, and M. K. McLaughlin. Pre-eclampsia: an endothelial cell disorder. Am. J. Ubstet. Gynecol. 161: 1200-1204, 1989. 22. Robertson, W. B., and P. J. Manning. Elastic tissue in uterine blood vessels. J. Puthd. 112: 237-243, 1974. 23. Rosenfeld, C. R. Distribution of cardiac output in ovine pregnancy. Am. J. Physiol. 232 (Heart Circ. Physiol. 1): H23LH235, 20.
1977. 24.
Rubanyi, G. M., and P. M. Vanhoutte. Nature of endotheliumderived relaxing factor: are there two relaxing mediators? Circ. Res. 61: 1161-1167,1987.
Soskel, N. T., T. B. Wolt, and L. B. Sandberg. Isolation and characterization of insoluble and soluble elastins. Methods Enzymol. 144: 196-214,1987. 26. Ueda, S., V. Fortune, B. S. Bull, G. J. Valenzuela, and L.
25.
CEREBRAL
ARTERIES
R143
D. Longo. Estrogen effects on plasma voIume, arterial blood pressure, interstitial space, plasma proteins, and blood viscosity in sheep. Am. J. Obstet. Gynecol. 155: 195-201, 1986. 27. Vane, J. R., E. E. Anggard, and R. M. Botting. Regulatory functions of the vascular endothelium. N. Engl. J. Med. 323: 2736, 1990.
28 Weiner, C. P., E. Martinez, D. H. Chestnut, and A. Ghodsi. ’ Effect of pregnancy on uterine and carotid artery response to norepinephrine, epinephrine, and phenylephrine in vessels with documented functional endothelium. Am. J. Obstet. Gynecol. 161: 2g
1605-1610,1989.
Weiner, C. P., E. Martinez, L. K. Zhu, A. Ghodsi, and D. H. Chestnut. In vitro release of endothelium-derived relaxing factor by acetylcholine is increased during the guinea pig pregnancy. Am. J. Obstet. Gynecol. 161: 1599-1605, 1989. 30. Zawadski, J. V., P. D. Cherry, and R. F. Furchgott. Comparison of endothelium-dependent relaxation of rabbit aorta by A23187 and by acetylcholine (Abstract). Pharmacologist 22: 271, .
1980.
Downloaded from www.physiology.org/journal/ajpregu by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 14, 2019.
D. HULL,
changes in ovine cerebral arteries DON
M. LONG,
LAWRENCE
D. LONGO,
AND
WILLIAM
J. PEARCE
Division of Perinatal Biology, Departments of Physiology and Obstetrics and Gynecology, Loma Linda University School of Medicine, Loma Linda, California 92350 Hull, Andrew D., Don M. Long, Lawrence D. Longo, and William J. Pearce. Pregnancy-induced changes in ovine cerebral arteries. Am. J. Physiol. 262 (Regulatory Integrative Comp. Physiol. 31): R137-R143, 1992.-We examined the effects of pregnancy on the ovine cerebral vasculature by comparing several characteristics of isolated endothelium-intact segments of three intracranial arteries including the middle cerebral (MCA), posterior communicating (PC), and basilar (BAS) arteries taken from pregnant sheep (138143 days gestation, term ~145 days) and nonpregnant controls. For comparison, segments of the extracranial common carotid (COM) artery were also studied. With pregnancy, vessel water content increased (5.458%) in all arteries except the PC. Additionally, cellular protein content increased in all arteries (4.4~50.0%). Arterial stiffness, as determined by passive stress-strain determinations, was significantly decreased during pregnancy in the MCA but not in the larger arteries. Maximum contractile responses, when normalized to vessel wall cross-sectional area, were consistently greater in arteries from pregnant than in those from nonpregnant animals (lo&49.7%). Relaxation to the endothelium-independent guanylate cyclase stimulator Snitroso-N-acetyl penicillamine (SNAP) increased with pregnancy only in the distal MCA (=17%). Endothelium-dependent relaxation to the calcium ionophore A23187 decreased only in the larger and more proximal COM (-39%). Thus pregnancy was associated with an increase in production of contractile force, a decrease in peripheral vascular stiffness, a decrease in the relaxant response to A23187 in the COM, and an increase in the relaxant response to SNAP in the MCA. Together, these findings indicate that pregnancy has widespread and important vessel specific cerebrovascular consequences that affect not only arterial composition, but also contractility and endothelial reactivity. cerebrovascular circulation; endothelium; vascular smooth muscle PREGNANCY induces major changes in the maternal cardiovascular system including an increase in blood volume of up to 45%, a doubling of cardiac output, a fall in
peripheral resistance, and major alterations in regional blood flow (11). Individual vascular beds, however, vary considerably in response to pregnancy (1, 23). Understandably, much attention has focused on the effects of pregnancy on the uteroplacental and renal circulations. However, examination of other vascular beds is important in determining
the more widespread
adaptations
of
the circulatory system to gestation. Elucidation of these changes should enhance our understanding of both normal and complicated pregnancy. One area worthy of examination during pregnancy is endothelial function. The endocrine and paracrine influences of the vascular endothelium are now widely recognized (27). The possibility that endothelial dysfunction is important in the pathogenesis of many disorders of pregnancy, including pre-eclampsia (21), makes the study of normal endothelial function and response during pregnancy of paramount importance.
In the present studies, we have examined adaptation of the cerebral circulation to pregnancy. Although the cerebral circulation has been held to be little affected by pregnancy (16, 23), relatively few studies of the effects of pregnancy on this vascular bed have been conducted. Given the possible relations between vessel composition, structure, and function, we have examined water and protein content, thickness and stiffness, and reactivity to both endothelium-dependent and endothelium-independent vasodilators. Recognizing that any alterations in function related to pregnancy may be vessel specific, we have examined arteries from four levels of the cerebrovascular tree. METHODS From pregnant sheep (n = 12) at 138-143 days gestation (term ~145 days) and from nonpregnant control animals (n = 48) we obtained complete segments of three intracranial arteries including basilar (BAS), posterior communicating (PC), and middle cerebral arteries (MCA). For comparison, segments of the extracranial common carotid (COM) artery were also studied. The nonpregnant animals were nulliparous young adults, confirmed by the United States Department of Agriculture to be disease free, obtained from a local abbatoir. After initial dissection and removal of excess adipose and connective tissue, we cut the vessels into individual rings 3 mm long (COM and PC) and 5 mm long (BAS and MCA). We determined vessel wall thicknesses using projection microscopy in fresh unfixed representative transverse sections of each artery as previously described (19). The anatomical integrity of the vascular endothelium was established in samples of arteries from representative animals using scanning electron microscopy, and in all cases the endothelium was found to be physically intact. In separate parallel functional experiments, responses to a submaximal dose (1 PM) of A23187, an endothelium-dependent vasodilator, were consistent and reproducible. Together, these data indicate that the endothelium was intact and functional in the arteries used in the present studies. Water and protein determinations. We determined relative vessel water and protein content as previously described (20). In brief, we weighed blotted arterial segments before and after 48 h desiccation at 50°C to determine water content and assayed vessel protein content using an extraction designed to exclude connective tissue and structural proteins (8, 25). We quantified protein using the Bradford Coomassie Brilliant Blue assay (5). Contractility experiments. We mounted each vessel segment on paired wires between a low compliance force transducer (Kulite BG-lo), and a post attached to a micrometer used to vary baseline tension. We equilibrated the arteries at 38.5OC for 30 min in a bicarbonate Krebs solution containing (in mM) 122 NaCl, 25.6 NaHCOs, 5.56 dextrose, 5.17 KCl, 2.49 MgSOI, 1.60 CaC12, 0.114 ascorbic acid, and 0.027 disodium EDTA, continuously bubbled with 95% Og-5% COZ. We obtained micrometer readings, and thereby diameter measurements, for each arterial segment under unstressed conditions (CO.1 g tension), and similarly at an optimum resting tension of 0.5 g
0363-6119/92 $2.00 Copyright 0 1992 the American Physiological Society
R137
Downloaded from www.physiology.org/journal/ajpregu by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 14, 2019.
R138
PREGNANCY
AND OVINE
for the cerebral vesselsand 1 g for the carotid arteries. Vessel diameterwascalculatedastwice the measureddistancebetween mounting wires divided by 7~ We calculated force per unit cross-sectionalarea,asdescribedpreviously (19), asthe product of maximum tension and the accelerationdue to gravity divided by cross-sectionalarea. Cross-sectionalarea is calculated as twice the product of vessellength and vesselwall thickness at optimal tension, which is derived by dividing the measured vessel diameter at rest by the measuredvessel diameter at optimum tension and multiplying this value by the vesselwall thickness measuredat rest. We first contracted the arteries by exposing them to an isotonic potassiumKrebs solution containing 122 mM K+ and 31 mM Na’ (the composition of this solution was identical to that of our “normal” sodium Krebs solution except that there was an equiosmolarexchangeof potassiumfor sodium). After peak tensionswere reached,we washedthe vesselswith normal sodiumKrebs solution and allowed them to return to baseline levels of tension for 30 min. We then induced a secondcontraction using a mixture of 10 PM serotonin and 20 PM histamine. Previous studies have shown this mixture to produce maximal contractions with greater stability than any other method of contraction in the ovine cerebral arteries (18). During all contractility experiments, we continuously digitized, normalized, and recordedcontractile tensions using an on-line computer. Determination of stress-strainrelations. We mounted vessel segmentsasdescribedabove, and obtained micrometer readings of vesseldiameter at near zero tension and at baselinetensions of 0.5 g (cerebral arteries) and 1 g (COM). These micrometer readingsenabledus to calculate true unstressedresting vessel diameter (D,) and vesseldiameter at baselinetension (&). We then carried out a seriesof potassiumdepolarization contractions at resting vesseldiameters varied over the range D/Db 0.5-1.7, and recorded the resultant maximum active tensions. From thesedata, we wereable to plot active stress/strain curves for each vessel,using true values of wall stress,thus allowing the determination of optimum resting diameter (Do& for each vessel.This measurementof DOptis conceptually equivalent to the measurementof optimal length, Lo, usedby other authors. Once DOpthad been determined, we froze the vesselswhile still mounted on their wires by immersion in liquid nitrogen to eliminate any active component of wall stress.After thawing, we generated a passive stress-strain curve, again by varying vesseldiameter over the rangeD/DOpt0.5-1.7 and recording the resultant passivetension at each diameter. Vasorelaxantstudies.We studied the relaxant responsesto two agents, A23187 and S-nitroso-N-acetyl penicillamine (SNAP). The calcium ionophore A23187 is a receptor-independent stimulator of endothelium-derived relaxant factor (EDRF) releasefrom the vascular endothelium (30). Its use therefore served as an indicator of the endothelium-evoked vasorelaxation through stimulation of guanylate cyclase and accumulation of guanosine 3’,5’-cyclic monophosphate (cGMP). In contrast, the nitrosothiol SNAP acts directly on the vascular smooth muscle independent of the endothelium. The rapid releaseof nitric oxide from SNAP producesmaximal stimulation of guanylate cyclase,maximal production of cGMP, and hencemaximal relaxation (12). We contracted the vesselswith serotonin and histamine as describedabove and, after establishment of stable tone, added either 1 PM A23187or 10 PM SNAP and recorded the resulting changesin tension producedby eachvessel.We determined the relaxant responsefor each vesselby integrating the area under each tension-time course curve. We expressedthe average of these values as a percentage of maximum tension, to give averagepercent relaxations for each vesseltype to each agent. All methodsand nroceduresusedin this studv were annroved
CEREBRAL
ARTERIES
by the animal care committee at our institution. Statistics. Throughout the text, values are reported asmeans t SE, unlessotherwise indicated. Where a single animal contributed more than one vesselsegmentto any particular protocol, the individual results were averaged to a single value before analysis of group statistics. A total of 416 nonpregnant and 174 pregnant artery segmentswere used.For all statistical comparisons either a l-way or a 2-way analysis of variance (ANOVA) was employed. Integrals of the tension-time curves were determined by Simpson’sapproximation. Comparisonsof ratios were made using a Behrens-Fischer analysis. All probabilities ~0.05 were regardedas significant. RESULTS
Vessel cellular protein and water content. In both the nonpregnant and pregnant animals, there was no significant difference in protein content between arteries (Fig.
1). Only in the COM was pregnancy significant
associated with -a
increase in protein content. Water content
was not significantly different between arteries in either group (Fig. 1). With pregnancy there was a significant increase of ~5.5% in all but the PC. Vessel wall thickness. With pregnancy, there was a decrease of 6.8 t 0.9% in the COM and slight increases in wall thickness in all the intracranial cerebral arteries in the pregnant animals (Table 1). However, none of these changes were statistically significant. Contractility. The maximal contractile responses to 122 mM potassium were significantly higher in the COM than in the smaller intracranial arteries in both nonpregnant (15.2 t 0.8 g) (P < 0.0001 ANOVA) and pregnant (18.5 t 1.8 g) animals (P < 0.0001 ANOVA). Between-artery differences were not significant in either group for any other artery (nonpregnant: BAS 3.0 t 0.2 g, PC 2.6 t 0.2 g, MCA 2.9 t 0.2 g; pregnant: BAS 2.9 t 0.3 g, PC 3.0 t 0.4 g, MCA 3.6 t 0.5 g). Pregnancy was H
Nonpregnant
N
Pregnant
*
COM
_ BAS
-
PC
- MCA
-I
Fig. 1. Pregnancy-induced changes in cellular protein and water content in ovine cerebral arteries. Protein contents were calculated as a percentage of vessel dry weight, and water contents were calculated as the percent differences between wet and dry weights. Both are shown here as means k SE for the numbers of animals indicated at the base of each bar. COM, common carotid artery; ‘BAS, basilar artery; PC, posterior communicating artery; MCA, middle cerebral artery, ‘P < 0.05 comnared with nonnreenant.
Downloaded from www.physiology.org/journal/ajpregu by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 14, 2019.
PREGNANCY
Table
AND OVINE
1. Vessel wall thickness Artery Common carotid
Posterior communicating
Basilar
Middle cerebral
Nonpregnant 935k17 (48) 112t2 (44) 139t5 (44) 123k2 (44) Pregnant 871_t39 (12) 113*5 (11) 144,tlO (12) 13027 (12) Vessel wall thicknesses are given above in microns and are expressed as means t SE for the numbers of animals given in parentheses. Changes with pregnancy were not significant for any artery type.
(Yn 2E Ow8 *c 0.6 3 04 5 0.2 0
0.6
E = 0.4 E 5 0.2 00
Carotid
Basilar
l
8Q) O 8 .-s! 5 0.8
a
Common
Pouterior
Communicating
Middle
Cerebral
0.8 0.6 0.4 0.2 0
Fig. 2. Pregnancy-induced changes in maximal cerebrovascular stress production. For each artery type, the maximum force produced by exposure to either 122 mM potassium, or 10 PM serotonin with 20 PM histamine, was normalized relative to vessel wall cross-sectional area. The resultant values of maximum active stress are given here as means and standard errors for the numbers of animals indicated at the base of each bar. §P < 0.05 compared with all other nonpregnant vessels; tP < 0.05 compared with all other pregnant vessels; *P < 0.05 compared with nonpregnant.
associated with a significant rise only in the COM (21.8 t 0.9%) (P < 0.05, ANOVA). Figure 2 shows maximum contractile force normalized relative to vessel wall cross-sectional area. In nonpregnant animals, maximal wall stress was lowest in the COM (0.31 t 0.01 x lo6 dyn/cm2) and increased in the more peripheral vessels: BAS (0.36 t 0.03 x lo6 dyn/ cm2), MCA (0.45 t 0.03 x lo6 dyn/cm2), and PC (0.55 t 0.04 x lo6 dyn/cm2). Wall stress was significantly higher in the PC compared with each of the other arteries. In the pregnant group, the PC again produced significantly more wall stress (0.80 $- 0.1 X lo6 dyn/cm’) than the other vessels, COM (0.43 t 0.02 x lo6 dyn/cm2), BAS (0.54 t 0.07 x lo6 dyn/cm2), and MCA (0.49 t 0.07 x lo6 dyn/cm2). Increases in maximal wall stress associated with pregnancy were significant in the COM (41.2 t 0.3%), BAS (49.7 t 0.6%), and PC (46.7 t 0.6%) but were not in the MCA (10.1 t 2.4%). When the maximal contractile response to a mixture of 10 PM serotonin and 20 PM histamine was expressed as a percentage of that induced by depolarization with 122 mM potassium, a similar pattern across the vascular tree was seen in vessels from both nonpregnant and pregnant animals. The greatest sensitivity to the agonists used was seen in the PC (nonpregnant 100.4 t 5.4%, pregnant 101.5 $- 12.8%) and the least in the COM
CEREBRAL
R139
ARTERIES
(nonpregnant 30.5 t 2.3%, pregnant 25 t 2.7%). The MCA (nonpregnant 68.7 t 5.6%, pregnant 69.6 t 7.2%) and BAS (nonpregnant 54.3 t 4.4%, pregnant 70.4 k 7.3%) showed intermediate values. Between-vessel differences were significant in all cases except in the pregnant BAS vs. MCA. None of the changes seen with pregnancy were statistically significant. Vessel diameters and passive wall stiffness. Not surprisingly, in both the nonpregnant and pregnant groups, resting vessel diameters were lower in the more periphera1 vessels (Fig. 3). None of the changes in resting diameter or optimum diameter seen with pregnancy were statistically significant (ANOVA). In the nonpregnant animals there were no significant differences in passive compliance between arteries (Fig. 4). However, in the pregnant group the MCA was significantly less stiff than the other vessels, which all had similar compliance values. Pregnancy produced a significant change in compliance only in the MCA. Vasorelaxant studies. The relaxant response to A23187 was similar in all four artery types obtained from the nonpregnant animals, ranging from 26.3 t 5.5% (MCA) to 37.6 t 6.5% (PC) (Fig. 5). Only the pregnancy-associated reduction in relaxation response to A23187 in the COM (from 33.2 t 3.5% to 20.1 t 2.5%) was significant. The degree of vasorelaxation produced by SNAP was significantly greater than that produced by A23187 in all vessels in both pregnant and nonpregnant groups. The smaller, more peripheral arteries relaxed significantly more (~75%) than the proximal COM (~40%). Only in the MCA was a significant effect of pregnancy seen on the ability of SNAP to relax the vessels (an increase from 64.9 t 2.2 to 75.9 t 2.4%). An estimate of the relative ability of the endothelium to recruit vasorelaxation was obtained by calculating the ratio of relaxation produced by A23187 to that produced n E E 5 5 E m E
Common
Carotid
Basilar
a
2
6
1.5
1,
’ 2 0 _
0.5
0
Posterior
Do
Communicating
Dopt.
Middle
Do
Cerebral
Dopt.
Nonpregnant
Pregnant RI Fig. 3. Pregnancy-induced changes in resting and optimum diameters in ovine cerebral vessels. Resting diameter at zero tension (Do) and optimum diameter (Do&, determined from active length-tension curves as described in the text, are given here for each vessel type. Values are means ,t SE for the number of animals given at the base of each bar. No significant differences between pregnant and nonpregnant animals were observed.
Downloaded from www.physiology.org/journal/ajpregu by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 14, 2019.
R140
PREGNANCY +
Common
Nonpregnant
6
AND OVINE
ARTERIES
s
Pregnant
Carotid
CEREBRAL
n
s 2
Basilar
s 5
N
Nonpregnant
Pregnant
60 1
a
v)
0.4
0.6
0.8
1
1.2
0.4
0.6
0.8
1
1.2
v) d) CI
Posterior
Communicating
Middle
Cerebral
1
n
.p 0.8 s E 0.6 0.4
0.6
0.8
1
1.2
0.4
0.6
0.8
1
1.2
L/Lo Fig. 4. Pregnancy-induced changes in passive stress-strain relations in ovine cerebral arteries. For each artery type, passive stress-strain curves are given for the range L/L,, 0.4-1.2, where L is true vessel diameter and L, is optimal vessel diameter. Values are given as means; errors bars are shown where larger than the symbols used. *P < 0.05 compared with all other nonpregnant and pregnant vessels; n = 6 for all but the carotid arteries from pregnant animals, where n = 4.
by SNAP (Fig. 5). There was a significant decrease in the ratio in the COM from 0.78 t 0.14 in the nonpregnant animals to 0.50 t 0.12 in the pregnant group. No consistent effect of pregnancy was seen in the smaller arteries. DISCUSSION
In humans, pregnancy has been shown to have little effect on global cerebral blood flow (16). This absence of effect has also been demonstrated in sheep by Rosenfeld (23) using radionuclide-labeled microsphere techniques. Overall, ovine cardiovascular adaptations to pregnancy are generally less pronounced than those seen in humans (26). Nonetheless, it is apparent from the present results that the ovine cerebral vasculature is significantly affected by pregnancy. It follows, therefore, that the vascular changes here described must be in balance with other circulatory changes associated with pregnancy to maintain constancy of cerebral perfusion. Total body water, in particular the extracellular fluid volume, is well known to increase in pregnancy. This increased state of hydration is shown in all but the PC (Fig. 1) with increases of =5.5% in vessel water content. Although their data were not statistically significant, Griendling et al. (10) found increases in water content of 2% in carotid and 7% in uterine arteries taken from pregnant sheep compared with nonpregnants. Because relative water content increases during pregnancy, it
a 3
0.4
Fi 2
0.2 0
COM
BAS
.
PC
. MCA
’
Fig. 5. Pregnancy-induced changes in relaxation response to A23187 and SNAP. For each artery type, percent relaxation produced by exposing previously contracted vessels to either 1 PM A23187 or 10 /IM SNAP is shown. Also given is the ratio of A23187-induced relaxations to SNAP-induced relaxations. The relaxation responses are shown as means k SE for the number of animals indicated at the base of each bar. COM, common carotid artery; BAS, basilar artery; PC, posterior communicating artery; MCA, middle cerebral artery. *P < 0.05 compared with nonpregnant.
follows that other vessel wall constituents may decrease. Although we have not specifically addressed this possibility, others (10) have found no significant change in ovine carotid artery fat or connective tissue content with pregnancy. Another possibility is that the mucopolysaccharide content is altered, as suggested by others in other vascular beds (22). Estrogen, derived from dehydroepiandrosterone sulfate produced by the fetal adrenal, stimulates the production of renin substrate and stimulates aldosterone secretion, thus encouraging sodium and water retention. Progesterone similarly increases renin activity and aldosterone levels. Thus the driving force for the changes in vessel hydration is likely to be hormonal. The lack of change in water content in the PC may reflect differences in numbers or level of activation of sex hormone receptors at different sites in the vasculature. The fraction of protein we assayed included only cellular and enzymatic proteins, and thus the data given in Fig. 1 do not reflect any pregnancy-related alterations in vascular connective tissue content. Minor increases in relative protein content were seen in the smaller arteries, but a significant rise of 50% was seen in the COM. As with our measurements of water content, the protein measurements were relative to total dry weight, implying
Downloaded from www.physiology.org/journal/ajpregu by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 14, 2019.
PREGNANCY
AND
OVINE
that some other cellular constituent is decreasing its relative contribution to dry mass. Annibale et al. (3), using a different assay technique, found an increase in vessel protein content with pregnancy in ovine uterine but not carotid or renal arteries. The reason for the difference between this and our findings may be that their values were expressed as percentages of wet weight, and the changes in vessel water content with pregnancy may vary in a site-specific manner. No significant alterations in vessel wall thickness were associated with pregnancy, nor were any clear patterns of variation discernable in the arteries studied. No attempt was made to quantify the relative contribution each component of the vessel wall made to overall thickness or changes therein. A reduction in both media and overall wall thickness in 165-pm mesenteric arteries (approximately half the diameter of our smallest artery, the MCA) was found in pregnant rats at term by other investigators (17). It may be that such structural alterations are limited more to resistance-sized vessels remote from the uteroplacental circulation. In contrast, the dramatic growth in the uterine arteries with pregnancy is well recognized. Several investigators have demonstrated in a variety of vessels in several different species that optimal resting tension increases with pregnancy (17, 28, 29). Figure 3 indicates that optimal vessel diameter, and by extension optimal resting tension, does not change uniformly across the vascular tree with pregnancy. However, in this study, none of the changes seen were statistically significant. The first published study of changes in arterial mechanics with pregnancy showed no effect on the passive compliance of rabbit aorta (13). Subsequent studies of ovine carotid artery have corroborated this finding (2, 10). In contrast, studies of resistance-sized vessels have shown significant pregnancy-associated decreases in passive stiffness (17). Our findings in the COM, BAS, and PC agree with the earlier studies conducted in the larger arteries; that is, we observed no significant change in passive compliance with pregnancy (Fig. 4). In contrast, and consistent with studies in smaller arteries, we observed a significant loss of stiffness in the MCA, which in our studies was x400 pm in diameter. Aside from these apparent modest changes in stiffness, the resting tone of the arteries may have also increased, as suggested by the observed values of D, (Fig. 3). These values, which also exhibited vessel-specific changes with pregnancy, decreased in the COM, BAS, and PC but remained unchanged in the MCA. Overall, there is good evidence that the general distensibility of the circulatory system increases in pregnancy (4). What then are the underlying changes in the arterial wall that can explain the observed changes in passive mechanical characteristics? Simple histological examinations, unfortunately, do not give the answer. Danforth (6) demonstrated fragmentation of the reticulum and attenuation of the elastica in rabbit aorta, but these changes appeared to have no effect on mechanical properties (13). Studies comparing estrogen-supplemented and untreated ovariectomized animals (7) suggested that estrogen stimulates collagen and elastin degradation at
CEREBRAL
ARTERIES
R141
differential rates in rat aorta, leading to a greater proportion of elastic fibers and hence greater distensibility. However, no absolute change in collagen, elastin, or collagen/elastin ratio was found with pregnancy in ovine carotid artery by Griendling et al. (lo), although a decrease in collagen was found in the uterine artery. The present data suggest that the changes taking place in the connective tissue jackets of arteries, and in resistance vessels in particular, are more subtle than thus far suggested. Alterations in the spatial relationship of the noncontractile elements of the vessel wall in a looser, more watery ground substance may occur. Altered collagen synthesis and turnover or modification of the crosslinkages between fibers may take place. These areas clearly require further investigation. Consistent with the observed increases in cellular protein, pregnancy was also associated with an increase in contractile capacity in all vessels, as determined by potassium depolarization. When these contractile tensions were normalized for vessel wall cross-sectional area (Fig. 2), significant increases were still seen in all but the MCA. Such changes may represent alterations in the coupling of membrane depolarization and Ca2+ entry into the vascular smooth muscle cells, and the subsequent activation of the contractile process. Clearly these complex pathways are potentially subject to modification by the hormonal influences of pregnancy. Comparisons between our findings and those of others are complicated by the fact that in most studies, contractile tensions are rarely normalized for wall cross-sectional area. McLaughlin and Keve (17) found a pregnancy-related decrease in contractile force in the rat mesenteric artery, a vessel considerably smaller than any of the arteries we studied. Consistent with the findings of McLaughlin and Keve, we observed a falloff in the size of the pregnancyinduced increase in force generation in the more peripheral vessels. Although Annibale et al. (2) demonstrated an increase in wall stress with pregnancy in the ovine uterine artery, their increase of ~40% in wall stress in the COM (cf. our finding of an increase of 41.2 t 0.3%) was not statistically significant. A later study (3) showed no increase in cellularity in the pregnant ovine uterine artery to account for the increase in stress production. There was, however, a demonstrable increase in both actin and myosin content as part of a general increase in total protein and a rise in the level of myosin light chain phosphorylation after contraction with phenylephrine. The ovine renal artery did not show these findings. Relaxation responses to A23187 were similar (~30%) in the smaller arteries in both pregnant and nonpregnant animals (Fig. 5). In the COM, however, there was a significant reduction in relaxation with pregnancy. In a recent study of acetylcholine-induced relaxation in guinea pig carotid and uterine arteries (29), pregnancy was associated with a significant enhancement of endothelium-mediated vasorelaxation, an effect attributed to an augmentation of EDRF activity. Support for this interpretation came from an earlier study of castrated estrogen-treated rabbits whose femoral artery relaxation response to acetylcholine was enhanced relative to that of untreated nonpregnants (9). However, additional experiments using A23187 in the same study showed no
Downloaded from www.physiology.org/journal/ajpregu by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 14, 2019.
R142
PREGNANCY
AND OVINE
enhancement of response by estrogen treatment. It seems likely, therefore, that the greater relaxation to acetylcholine seen in both groups of experiments is an estrogenmediated receptor effect rather than a fundamental alteration in EDRF production or release. It remains possible, however, that acetylcholine and A23187 may not release the same EDRF (24). In the present data, the equivalency of the responses to A23187 in the smaller arteries of both groups suggests that EDRF production in these arteries is unaltered by pregnancy. In contrast with this finding, pregnancy was associated with a decrease in A23187-induced relaxation in the COM. The reason and mechanism for this depression remain unclear. Basal EDRF production may be increased in pregnancy, thereby reducing the apparent “relaxing capacity” of the endothelium in response to agonists. This is supported by the finding of an increase in basal cGMP levels in the pregnant ovine uterine artery (15). Equally, changes in any of a vast number of steps could explain this finding, including a decrease in the calcium sensitivity of endothelial cells, a decrease capacity for EDRF synthesis, or a limitation of EDRF precursor, to name only a few possibilities. One mechanism that cannot explain the decreased sensitivity to A23187 in the pregnant COM is that the vascular guanylate cyclase, which mediates the effects of EDRF, was somehow less sensitive to EDRF in the pregnant animals. SNAP, which releases nitric oxide and thus produces relaxation via the same pathway as that used by the endothelium (l4), produced rapid and maximal relaxation in all arteries (Fig. 5). Responses to SNAP were unaffected by pregnancy, except in the most distal artery studied, the MCA, where an enhancement was observed. Thus a decrease in sensitivity to EDRF cannot explain the decreased responsiveness to A23187 observed in the pregnant COM. Instead, a direct effect of pregnancy on endothelial synthesis and/or release of EDRF is suggested. In that endothelial dysfunction has been suggested as a possible factor in pre-eclampsia (21)) this latter suggestion warrants additional study. Conclusions. The present study demonstrates that pregnancy affects the cerebral vascular bed, despite its remoteness from the conceptus. Although increases in water, protein content, and force-generating ability were seen as a general phenomenon, other parameters changed in a highly vessel-specific manner. The passive mechanical characteristics of the larger arteries were unchanged, whereas the distal MCA showed a significant loss of stiffness. The capacity of the endothelium to elicit vasorelaxation was altered with pregnancy only in the COM, where a decrease was demonstrated. The maximal relaxation potential of the vascular smooth muscle changed only in the MCA, which showed an increase. Alterations in both the passive and active characteristics of the MCA may parallel the general decrease in peripheral vascular resistance seen elsewhere with pregnancy. Endothelial function is altered only in the COM, although the reason for this is uncertain. Clearly, the results of the present study, when combined with the work of others, indicate that pregnancy has profound effects on the vascular system. Because many of these effects are vessel and site specific, findings in one bed
CEREBRAL
ARTERIES
may not be directly extrapolated
to another.
The authors extend their appreciation to Charles White and Al van Varik for their excellent technical assistance with these studies. The authors also thank Dr. Louis Ignarro for the SNAP used in these studies. This project was supported by National Institutes of Health Grants HL-41347 and HD-03807 and Grant 90-121 from the California Affiliate of the American Heart Association. Address for reprint requests: A. D. Hull, Division of Perinatal Biology, Loma Linda University School of Medicine, Loma Linda, CA 92350. Received 25 February 1991; accepted in final form 9 August 1991. REFERENCES 1. Abramson, D. I., K. Flachs, and S. M. Fierst. Peripheral blood flow during gestation. Am. J. Obstet. Gynecol. 45: 666-671, 1943. 2. Annibale, D. J., C. R. Rosenfeld, and K. K. Kamm. Alterations in vascular smooth muscle contractility during ovine pregnancy. Am. J. Physiol. 256 (Heart Circ. Physiol. 25): H1282-H1288, 1989. 3. Annibale, Kamm.
D.
J.,
C. R.
Rosenfeld,
J.
T.
Stull,
and
K.
K.
Protein content and myosin light chain phosphorylation in uterine arteries during pregnancy. Am. J. Physiol. 259 (Cell
Physiol. 4. Bonnar,
28): C484-C489, 1990. J., I. MacGillivray, and E. M. Symonds. reactivity in pregnancy. In: Pregnancy Hypertension.
Vascular Baltimore,
MD: University Park Press, 1980. M. M. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72: 248-254, 1976.
5. Bradford,
6. Danforth,
D. N., P. Manola-Estrella,
and J. C. Buckingham.
The effect of pregnancy and of Enovid on the rabbit vasculature. Am. J. Obstet. Gynecol. 88: 952-962, 1964. 7. Fischer, G. M. In vivo effects of estradiol on collagen and elastin dynamics in rat aorta. Endocrinology 91: 1227-1232,1972. 8. Furoto, D. K., and E. J. Miller. Isolation and characterization of collagens and procollagens. Methods Enzymol. 144: 41-61,1987. 9. G&lard, V., V. M. Miller, and P. M. Vanhoutte. Effect of 17
beta-estradiol 10.
J. Pharmacol. Griendling,
on endothelium-dependent
responses in the rabbit.
Exp. Ther. 244: 19-22, 1988. K. K., E. 0. Fuller, and R. H. Cox.
Pregnancyand carotid arteries. Am. J. 17): H658-H665, 1985. Adaptations of metabolism
induced changes in sheep uterine Physiol.
248 (Heart Circ. Physiol. and G. Chamberlain.
11. Hytten, F., and nutrient Boston, MA: 12. Ignarro, L. A. L. Hyman,
requirements. In: Clinical Physiology Blackwell, 1980, p. 345-350. J.,
H. Lippton, P. J. Kadowitz,
in Obstetrics.
J. C. Edwards, W. H. Baricos, and C. A. Gruetter. Mechanism
of vascular smooth muscle relaxation by organic nitrates, nitrites, nitroprusside and nitric oxide: evidence for the involvement of Snitrosothiols as active intermediates. J. Pharmacol. Exp. Ther. 218: 739-749,198l. 13. Johnson, W. L., J. T. Conrad,
D. Whitney,
and N. Graybeal.
The effect of pregnancy and norethynodrel with mestranol (EnOvid) on length-tension relationship in the rabbit aorta. Am. J. Obstet. Gynecol. 93: 179-180, 1965. 14. Lincoln, T. M. Cyclic GMP and Pharmacol. Ther. 41: 479-502,1989. 15. Magness, R. R., C. R. Rosenfeld,
mechanisms of vasodilation.
and P. W. Shaul. Role of endothelium in the control of CAMP and cGMP production by uterine artery vascular smooth muscle in ovine pregnancy (Abstract). Proc. Sot. Gynecol. Invest. 265, 1991. 16. McCall, M. L. Cerebral blood flow and metabolism in toxaemias of pregnancy. Surg. Gynecol. Obstet. 89: 715-721, 1949. 17. McLaughlin, M. K., and T. M. Keve. Pregnancy-induced changes in resistance blood vessels. Am. J. Obstet. Gynecol. 155: 1296-1299,1986. 18. Pearce, W. J., and
S. Ashwal. Developmental changes in thickness, contractility, and hypoxic sensitivity of newborn lamb cerebral arteries. Pediatr. Res. 22: 192-196, 1987. 19. Pearce, W. J., S. Ashwal, and J. Cuevas. Direct effects of graded hypoxia on intact and denuded rabbit cranial arteries. Am.
Downloaded from www.physiology.org/journal/ajpregu by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 14, 2019.
PREGNANCY
AND OVINE
J. Physiol. 257 (Heart Circ. Physiol. 26): H824-H833, 1989. Pearce, W. J., A. D. Hull, D. M. Long, and L. D. Longo. Developmental changes in ovine cerebral artery composition and reactivity. Am. J. Physiol. 261 (Regulatory Integrative Comp. Physiol. 30): R458-R465, 1991. 21. Roberts, J. M., R. N. Taylor, T. J. Musci, G. M. Rodgers, C. A. Hubel, and M. K. McLaughlin. Pre-eclampsia: an endothelial cell disorder. Am. J. Ubstet. Gynecol. 161: 1200-1204, 1989. 22. Robertson, W. B., and P. J. Manning. Elastic tissue in uterine blood vessels. J. Puthd. 112: 237-243, 1974. 23. Rosenfeld, C. R. Distribution of cardiac output in ovine pregnancy. Am. J. Physiol. 232 (Heart Circ. Physiol. 1): H23LH235, 20.
1977. 24.
Rubanyi, G. M., and P. M. Vanhoutte. Nature of endotheliumderived relaxing factor: are there two relaxing mediators? Circ. Res. 61: 1161-1167,1987.
Soskel, N. T., T. B. Wolt, and L. B. Sandberg. Isolation and characterization of insoluble and soluble elastins. Methods Enzymol. 144: 196-214,1987. 26. Ueda, S., V. Fortune, B. S. Bull, G. J. Valenzuela, and L.
25.
CEREBRAL
ARTERIES
R143
D. Longo. Estrogen effects on plasma voIume, arterial blood pressure, interstitial space, plasma proteins, and blood viscosity in sheep. Am. J. Obstet. Gynecol. 155: 195-201, 1986. 27. Vane, J. R., E. E. Anggard, and R. M. Botting. Regulatory functions of the vascular endothelium. N. Engl. J. Med. 323: 2736, 1990.
28 Weiner, C. P., E. Martinez, D. H. Chestnut, and A. Ghodsi. ’ Effect of pregnancy on uterine and carotid artery response to norepinephrine, epinephrine, and phenylephrine in vessels with documented functional endothelium. Am. J. Obstet. Gynecol. 161: 2g
1605-1610,1989.
Weiner, C. P., E. Martinez, L. K. Zhu, A. Ghodsi, and D. H. Chestnut. In vitro release of endothelium-derived relaxing factor by acetylcholine is increased during the guinea pig pregnancy. Am. J. Obstet. Gynecol. 161: 1599-1605, 1989. 30. Zawadski, J. V., P. D. Cherry, and R. F. Furchgott. Comparison of endothelium-dependent relaxation of rabbit aorta by A23187 and by acetylcholine (Abstract). Pharmacologist 22: 271, .
1980.
Downloaded from www.physiology.org/journal/ajpregu by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 14, 2019.
