- Email: [email protected]
Pharmacodynamic study of maturation and closure of human umbilical arteries
Pharmacodynamic study of maturation and closure of human umbilical arteries Richard P. White, PhD Memphis, Tennessee The contractile effects of 19 factors on isolated human arterial segments at term pregnancy were quantified, and 14 contractile agents were similarly applied to preterm (23 to 35 weeks) umbilical arteries. Responses to potassium chloride were used to normalize the data. At comparison with the term vessel, the preterm artery contracted more to angiotensin II and arachidonic acid and was more sensitive to oxytoxin. Contractions were greater in term arteries to vasopressin, norepinephrine, prostaglandin O2 , and prostaglandin E2 but similar in both group of arteries to bradykinin, histamine, acetylcholine, and prostaglandin F2a • Neuropeptide Y, linoleic acid, uridine triphosphate, and thrombin were ineffective. Hyperoxia inconsistently induced weak, short-lived contractions. Contractions to cooling manifested marked desensitization and tachyphylaxis. Serotonin was the only agonist that displayed the pharmacodynamic features most likely to be important for closure: potency, efficacy, and long duration of action (>2.5 hours). It was postulated that cellular elements surrounding umbilical vessels are primary sources of vasoactive agents that are important to closure of the fetoplacental circulation at birth. (AM J OBSTET GYNECOL 1989;160:229-37.)
Key words: Umbilical arteries, pharmacology, closure phenomenon
It is characteristic among mammals for the umbilical cord to remain attached to the placenta for prolonged periods after birth. Closure of the extracorporeal circulation is therefore essential for survival, and the mechanism responsible for closure must operate for indefinite periods. Histologic and blood flow studies of humans indicate that arterial constriction is more important than narrowing of the umbilical veins to closure, that vasoconstriction peaks at about 45 seconds, and that flow decreases most at about 120 seconds after birth.1. 2 However, flow continues for at least several minutes beyond this point, during which time the placenta reportedly begins to detach from the uterus. 2 Previous studies concerned with the placental circulation have identified a number of diverse physiologic stimuli that might contribute to, or elicit, closure of the extracorporeal circulation at birth. The stimuli studied include eicosanoids, norepinephrine, epinephrine, acetylcholine, serotonin, histamine, bradykinin, angiotensin II, oxytocin, vasopressin, hemoglobin, hyperoxia, and temperature.'" Certain prostaglandins, serotonin, histamine, and bradykinin appear to be the most effective contractile agents, although bradykinin From the Department of Pharmacology and the Newborn Center, Untversity of Tennessee Medical Center. Supported by National Institutes of Health, United States Public Health Service, grant NS 21405. Received for publicatIOn January 8, 1988; revised May 5. 1988; accepted June 13, 1988. Repnnt requests: RIchard P. White, PhD, Department of Phannacology, UniverSIty of Tennessee Medical Center, Memphis, TN 38163.
may induce tachyphylaxis. s However, the identification of factors responsible for closure has been largely based on acute responses of umbilical vessels to stimuli, the responses often were not quantified, and the relevancy of the duration of the response to closure was not evaluated. Since increasing oxygen tension may elicit contraction of umbilical vessels, some investigators have emphasized the role oxygen might play in closure. The results are equivocal in that the vasoconstriction produced by oxygen is reportedly inconsistent and not sustained. s, 1O Although cooling umbilical vessels appears to reliably provoke constriction," 5. 10 the duration of this response has not been studied, and mammalian births often occur in hot environs. Many vasoactive agents have been suspected to playa role in closure, but the role played by desensitization and tachyphylaxis in reducing the effectiveness of those agents has not been systematically explored. The present pharmacodynamic study was performed to characterize the contractile responses of mature umbilical arteries to a wide range of stimuli, with special attention paid to the phenomena of desensitization and tachyphylaxis. Contractile responses produced by umbilical arteries of preterm infants also were studied to ascertain whether the pharmacodynamic property of the artery changes significantly with maturation. Material and methods
Handling of the arteries. The umbilical arteries were considered mature if the infant was ;;;.38 weeks' ges-
229
230 White
January 1989 Am J Obstet Gynecol
5
-S
n =20
Term
4
*
.1----1-----1 Preterm
oc:: 3
.~
.=02 c::
8
1
*
,, ,, I
* ,,
, ,,
I
I
I
1
o~~~~--~--~
K+ 10
30
50
90 mM
Fig. 1. Dose-response curves comparing responses of preterm and term umbilical arteries to potassium chloride. Astensks indicate significant difference between corresponding points (P < 0.05).
tation. The outside diameter of 14 such arteries was 2.23 ± 0.08 mm. Premature arteries were obtained from 27 infants who were .:;;35 weeks' gestation. The average gestational length was 29.8 ± 1.1 weeks, and the infants weighed 1472 ± 142 gm (390 to 2540 gm). The outside diameter of the preterm arteries was 1.26 ± 0.16 mm. The vessels were routinely extirpated from the mid portion of the cord, except for a few experiments performed on umbilical arteries located within 5 cm of the abdominal wall. The lumen and outside of the arteries were washed with a physiologic salt solution and cleaned of superfluous material. A 4 mm segment (ring) was cut and slipped onto two parallel prongs of a tissue chamber. One prong was fixed and the other movable, being attached to a transducer for recording the tension of the arterial segment. The chamber was filled with 10 ml of the physiologic salt solution having the following composition (in millimoles per liter): sodium chloride, 118; potassium chloride, 4.7; magnesium sulfate, 1.2; sodium bicarbonate, 25; potassium monophosphate, monobasic, 1.2; calcium chloride, 2.5; glucose, 11.0; the pH was adjusted to 7.4 by adding hydrochloride. A mixture of 95% oxygen and 5% carbon dioxide was used to routinely aerate the tissue bath and a buffer reservoir. The bath temperature was 37° C. The arterial segment was allowed to incubate for 1.5 hours, during which time it is washed with physiologic salt solution about every 20 minutes. This wash is an irrigation from the bottom of the tissue chamber to an overflow so that the arterial ring is not exposed to air during washout. Arterial tension was initially set at 2 gm by means of a fine-positioning device (FTA 1011, Hewlett-Packard). Additional tension was applied as needed at about 20minute intervals in order to set a steady-state resting
tone of 0.5 to 1.0 gm. One hour later the responsiveness and the stability of the response for each artery were tested at least three times with 10, 30, 50, and 90 mmollL potassium chloride until a stable response was achieved. Two tissue baths were used so that arterial segments from the same individual would be studied in duplicate . The type of experiment performed on each ring differed so that any unknown peculiarity of an artery from one individual would not bias the overall findings. Drugs. The following drugs were obtained from Sigma Chemical Co., St. Louis: acetylcholine, angiotensin II, arginine vasopressin, bradykinin, histamine, neuropeptide Y, norepinephrine, oxytocin, phenylephrine, serotonin creatine sulfate, thrombin, uridine triphosphate, sodium arachidonate, linoleic acid, and prostaglandin D2 , E 2 , and F2u • These drugs were solubilized in distilled water, except that thrombin was dissolved in 0.9% sodium chloride, linoleic acid in water to which IN sodium hydroxide was carefully added, and neuropeptide Y in physiologic salt solution. The serotonin creatine sulfate antagonist cinanserin used in one study was obtained from E. R. Squibb and Sons, Inc., Princeton, N. J. The maximum volume of the solutions added to the bath per experiment was 0.2 m!. The various vehicles used to dissolve the drugs were inactive. Basic experimental design. The agonists were applied to the bath in increasing concentrations to give dose-response data. The concentrations applied ranged from 10- 7 to 10- 4 mollL, except for thrombin, which was applied as 0.1,1.0, and 10 NIH units per milliliter. The magnitude of the response was observed for 15 minutes, and its decay during that period was used as an index of desensitization. To determine whether the desensitization phenomenon persisted after the agonist was removed, the vessel was washed several times, and a period of 20 to 30 minutes elapsed to permit the artery to return to its original state. Then the agonist was applied again and the second response observed for 15 minutes. If the second response was statistically less than the first, the artery was considered to have manifested tachyphylaxis to the agent. Arteries that manifested complete tachyphylaxis to one agent, or failed to respond to a particular agonist, were exposed in addition to 90 mmoll L potassium chloride to ascertain whether the contractile mechanism was still operative. The maximum response to potassium chloride was used as a standard for comparing responses elicited by other agonists because it is independent of receptors and because of the great variation in contractile force generated among isolated arteries. Responses to the agonists were expressed as mean ± SEM gram force or percent of maximum response to potassium chlo-
Umbilical arteries
Volume 160 Number I
T
~ft
~
~ E
1 1 0
i9 II
.'
!Q
c
:~
T
0
J
e
50
1 1 1 1 1 1
P 1 1 1 1 •
6~
it 1
~P 1 1 1 1 1 1 0
e
6 P
,!
· ~ d T
II II II
\'
1 1 1 1
1 1 1
II
1
0
6 1 1
po
1
'~
• •o o
P - Preterm Arteries T - Term Arteries
J
1
pi
1 1 1
: P
~~ 1 1 1
6
1 1
•
10-7 M
1 1 1 1
J
":6 •
10-6M
J
,
•
Sundry
Av. of Max Response Response at 10"M 10-5M
T
P
"J 1 "J 1
o
.
T
J J
:::!1 0
~ 8
J
9' : II 1
LI>
100 1 + I> ::.::
U
Fatty Acids
Amlnes
Pepbdes
150
231
~ 1*
1 1 1 1 1 1
~~ 1 ri1T
c:
15
E
e
~
PT PT
Fig. 2. Summary of acute effects of agonists. BK, Bradykinin; ANG, angiotensin II; OXY, oxytocin; AVP, arginine vasopressin; NPY, neuropeptide Y; Hist, histamine; 5-HT, serotonin creatine sulfate; ACh, acetylcholine; NE, norepinephrine; PGF2., prostaglandin F2.; PGE" prostaglandin E2 ; PGD2 , prostaglandin D2 ; AA, arachidonic acid; LA, linoleic acid; UTP, uridine triphosphate. Asterisks indicate significant differences between average maximum response obtained in preterm and term umbilical arteries. Highest contraction and highest concentration shown may not always correspond because occasionally individual vessels responded best to a previous concentration. Except for NPY (n = 3), each bar graph represents 5 to 13 experiments. For clarity all SEM were omiu!'!d. Concentration of thrombin (0.1 to IOU / ml) is not represented in legend.
ride. The level of significance among the different responses was assessed by the appropriate Student t test.
Results Comparison of magnitude of responses by preterm and term arteries to agonists. Fig. 1 illustrates that the force (grams) of contraction elicited by potassium chloride was on average significantly greater in the term artery than in the pre term vessel. Although it is evident from the standard errors shown that the magnitude of the responses varied considerably from one individual to another, the arterial reaction to potassium chloride in each was remarkably stable and did not exhibit tachyphylaxis. Fig. 2 summarizes the responses elicited by agonists in preterm and term umbilical arteries. It is evident that the maximum contractile response elicited by bradykinin, oxytocin, histamine, serotonin, acetylcholine, and prostaglandin F2• were comparable in both groups of arteries when normalized to potassium chloride. However, arginine vasopressin, norepinephrine, prostaglandin E" and prostaglandin D2 produced significantly (p < 0.05) greater effects in the mature vessel than in the preterm one. On the other hand, angiotensin II and arachidonic acid produced significantly greater contractile responses in the preterm artery than in the mature one.
It should also be noted that at 10- 6 mol/L the preterm vessel was more reactive than the term artery to both angiotensin II and oxytocin. Increasing the concentration tenfold produced no further response to angiotensin II but increased the response to oxytocin significantly in the term artery (Fig. 2). In general, the response produced by 12 of the agonists at 10- 6 mollL was relatively weak, being only "",50% of that obtained with potassium chloride (Fig. 2). It is further evident that neuropeptide Y, uridine triphosphate, linoleic acid, and thrombin were not contractile agents in umbilical arteries. In contrast, bradykinin, histamine, and serotonin creatine sulfate produced responses that were> 50% of the potassium chloride induced response at 10- 6 mollL. Decay of response with time and tachyphylaxis. In these experiments the decrease in the maximal contractile response that occurred over a I5-minute interval was used as an index of desensitization. After washout and time allowed for tone to return to precontracted levels, the agonists were applied a second time. Any decrement in the second response to the agonist over the first was taken as an index of tachyphylaxis. Table I summarizes the findings. The responses produced by all of the peptides (Table I) decayed significantly and most induced tachyphylaxis. Moreover, the decay of the second response was
232 White
January 1989 Am J Obstet Gynecol
Table I. Comparison of maximum response, desensitization, and tachyphylaxis (recovery) elicited by agonists in preterm and term umbilical arteries* Response as % maximum response to potassium chlonde Agonist
Peptides Bradykinin Angiotensin II Oxytocin Arginine vasopressin Vasoactive amines 5-Hydroxytryptamine Histamine Acetylcholine Norepinephrine Fatty acids Prostaglandin F2a Prostaglandin
E2
Prostaglandin
D2
Arachidonic acid
Type of artery
Maximum
Preterm Term Preterm Term Preterm Term Preterm Term
113.8 107.2 31.4 1.6 28.8 23.8 23.6 49.5
Preterm Term Preterm Term Preterm Term Preterm Term
134.9 136.8 133.9 97.6 44.6 56.3
Preterm Term Preterm Term Preterm Term Preterm Term
106.6 118.8 71.2 119.1 63.4 108.4 49.9 4.3
± ± ± ± ± ± ± ±
1
At 15 min
21 25 17 0.8 9 7 6 18
28.3 42.6 7.0 0.2 5.0 4.4 6.4 4.2
± 18 ± 16 ± 13 ± 15 ± 14 ± 17 0.0 39.7 ± 16
94.5 118.0 51.7 42.9 5.8 17.5
± ± ± ± ± ± ± ±
19 22 16 21 21 17 24 2
± ± ± ± ± ± ± ±
7* 6* 3* 0.4 2* 2* 3* 0.6*
± 22* ± 13 ± 17* ± 14* ± 3* ± 9* 0.0 3.8 ± 2*
68.5 64.2 27.7 36.7 20.9 68.4 4.2 0.8
± ± ± ± ± ± ± ±
18 24 12* 9* 6* 26 3* 0.6
Second maxImum response as % potassium chloride
39.5 ± 14* 103.9 ± 22 26.4 ± 13 0.0* 3.8 ± 0.7* 0.5 ± 0.4* 11.4 ± 4* 32.3 ± 9 131.6 129.9 63.6 82.1 18.3 32.0
± 24 ± 19 ± 17* ± 10 ± 13* ± 16* 0.0 8.8 ± 5.6*
109.1 ± 17 113.6 ± 22 31.1 ± 20* 130.9 ± 14 34.1 ± 15* 112.7±19 0.8 ± 0.9* 0.0
*AII responses expressed as percent of maximum response to potassium chloride; second response obtained after washout of first and basal tone had returned to precontracted state. Asterisks indicate significant (p<0.05) decay in response after 15 minutes or significant difference in magnitude of first response over second. Each response was based on six to nine experiments. Agonists ranked by class of drug and by efficacy in preterm arteries.
more accelerated than the first, suggesting that tachyphylaxis is an extension of the desenitization phenomenon (data not shown). Fig. 3, B, illustrates the marked desensitization and tachyphylaxis produced by bradykinin in one preterm artery. All vasoactive amines (Table I) except serotonin creatine sulfate significantly induced desensitization. In addition, serotonin creatine sulfate was the only agonist whose effect persisted after washout, a phenomenon illustrated in Fig. 3, C. The persistent contraction after washout was evidently caused by serotonin creatine sulfate because it was ultimately removed by frequent washings and because it could be completely reversed by 10- 5 mollL cinanserin. The marked desensitization and tachyphylaxis produced by norepinephrine in the term artery was replicated with phenylephrine. Phenylephrine was studied because it reportedly produces more prolonged responses than norepinephrine in most vessels and the type of a-adrenergic receptor agonist might affect the induction of tachyphylaxis. Desensitization and tachyphylaxis to phenylephrine were nearly total (data not shown).
Among the fatty acids (Table I), prostaglandin F2a clearly produced the best response in the preterm vessel and percentage-wise induced less desensitization than the other lipids (p < 0.05 compared at 15 minutes). Desensitization to sodium arachidonate was the most rapid. Preterm arteries, but not term vessels, became tachyphylactic to prostaglandins E2 and D2 • In the term artery the prostaglandins were essentially equal in efficacy, although sodium arachidonate was ineffective. In general maturation enhanced the effectiveness of prostaglandins whereas it paradoxically reduced that of arachidonate acid. Because serotonin creatine sulfate was the only agonist not easily removed by washout and it failed to induce desensitization or tachyphylaxis, another protocol was adopted to study serotonin creatine sulfate, in which the effect of either 10- 6 (n = 6) or 10- 4 (n = 6) mollL was observed for 2.5 hours in term arteries. Contrary to expectation, desensitization to 10- 6 mollL was greater than that of 10- 4 mollL (Fig. 4). The desensitization to 10- 0 mollL was not due to a refractoriness because additional serotonin creatine sulfate pro-
Umbilical arteries 233
Volume 160 Number I
A.
Umbilical Arteries
K+
W
(:]~ • •••
B.
PT
I~
10 305090mM
5MIN
BK 1st Trial
(;] ;----• • • • -8
-7
-5(logM)
-6
2nd Trial
PT
W
~-
,. W
-8•
c.
• -6•
•
-5 (log M)
-7
5-HT
~]
• -7
~
• log M) • -5• -4(
-6
W
W
T
~ -20 MIN-
Fig_ 3. Recordings to illustrate different responses to agonists. Note sustained tonic phase of reaction to potassium chloride (A), which is readily removed with washout (W). Desensitization (BK 1st Trial) and tachyphylaxis (2nd Tnal) to contractile effects of bradykinin are clearly evident in preterm (PT) vessel. Note that sustained contraction to serotonin creatine sulfate (5-HT) was difficult to remove by washout. The 20-minute marker applies only to 5-HT.
duced a marked response (Fig. 4). Also, the more rapid decline obtained with the lower concentration was apparently not due to oxidation of serotonin creatine sulfate in the bath because lO-3 mollL ascorbic acid provided no protection against the desensitization phenomenon (n = 4). Cooling nine term umbilical arteries suddenly from 37.5" C to an average of 19.4 ± 0.6° C produced a contraction that was 48.2% ± 5.1 % of the maximum elicited by potassium chloride (5.1 ± 0.9 gm). The cooling was achieved by a wash with 8° C buffer and maintained for 10 minutes by use of a thermistor as a guide. The contraction, however, was not maintained and ceased between 5 and 10 minutes. The vessel was then warmed to 25° C and again cooled to see if less change in temperature might produce less of a response. Although the contraction was less, repeating this last procedure several times produced progressively less effect (Fig. 5). Thus cooling produced both desensitization and tachyphylaxis in these arteries (Fig. 5). Cooling the arteries slowly (15 minutes) from 37.5° to 17° C by reducing the temperature of the water jacket failed to elicit a response. Moreover, the cooled arteries were not more sensitive or responsive to serotonin creatine sulfate in concentrations of lO-8 to lO-5 mollL (n = 6).
Contractions produced by elevating oxygen content oftissue bath. In these experiments the maximum contraction to potassium chloride was established, and later
the concentration of oxygen aerating the artery was reduced from 95% to 0% (5% carbon dioxide in nitrogen), 2%, or 8% for 1.5 hour and then increased suddenly to 12% or 95%. As seen in Table II, 12 of 33 arteries responded to hyperoxia (95% oxygen), but none responded when the concentration was changed from 8% to 12%. These later concentrations provided the vessel with near physiologic levels of oxygen with bath P02 concentrations of about 63 and 95 mm Hg, respectively, and represent percentagewise the least change in oxygen tension. The contractions elicited by higher changes in oxygen content were, nevertheless, inconsistent, relatively weak, and of short duration (Table II). Incidence of responding arteries to agonists. The agonists that elicited a response in each preterm artery tested were limited to bradykinin (11 of 11), acetylcholine (8 of 8), histamine (8 of 8), serotonin creatine sulfate (13 of 13), prostaglandin E2 (9 of 9), and prostaglandin F2u (9 of9). However, the incidence of responders to angiotensin II (9 of 11), oxytocin (11 of 12), arginine vasopressin (6 of 7), sodium arachidonate (4 of 5), and prostaglandin D2 (8 of 9) was high. No response occurred with norepinephrine (0 of 7), neuropeptide Y (0 of 3), uridine triphosphate (0 of 5), or thrombin (0 of 7). The agonists that produced a response in all term vessels tested were bradykinin (35 of 35), arginine vasopressin (lO of lO), acetylcholine (lO of lO), histamine
234
White
January 1989 Am J Obstet Gynecol
i
Bath ·C 19.4
17.8
16.8
17.0
2
3
4
50
~ "#. 40
:I
:!i! 30
10"M-(6) 10" M .----.(10)
10""
aft.r
150 min
O~~--~--~--~--~--~ 5
15
30
60
150
Decay with Tlma (min)
Fig. 4. Time course of tonic phase of contraction induced by 10- 6 or 10- 4 mollL serotonin. Note that arteries exposed to 10- 6 mollL (5-HT) manifested the greatest desensitization but nevertheless responded maximally to 10- 4 mollL at 150 minutes. Asterisks represent significant differences between corresponding points (p < 0.05). Numbers in parentheses are number of experiments.
(13 of 13), norepinephrine (11 of 11), phenylephrine (12 of 12), serotonin creatine sulfate (21 of 21), prostaglandin D2 (8 of 8), prostaglandin E2 (11 of 11), and prostaglandin F2a (20 of 20). Those that produced no or a low incidence of response were angiotensin II (4 of lO), sodium arachidonate (l in 5), linoleic acid (1 of 4), thrombin (0 of 9), and uridine triphosphate (0 of 7). The capacity of angiotensin II to elicit a higher incidence of response in the preterm artery (81.8%) than in the term one (40%) was unexpected as were the effects obtained with sodium arachidonate. The fact that many agonists produced responses in all arteries tested suggests that, in the absence of further analysis, these might be considered important to the closure phenomenon. Preliminary experiments performed on umbilical arterial segments from proximal portion of cord. To ascertain whether segments of the umbilical artery near the abdominal wall might differ pharmacodynamically from those located midway in the cord, arterial rings were excised from the proximal portion of the cord, within 5 cm of the abdomen. These cords were obtained from eight preterm and four term infants. The following compounds were applied to at least two arteries from different individuals: bradykinin, oxytocin, angiotensin II, arginine vasopressin, histamine, norepinephrine, serotonin creatine sulfate, prostaglandins D2 and E2 , and thrombin. Since the responses produced did not differ in any obvious manner from those obtained from arterial segments more distal in the cord, these experiments were discontinued.
8
.s
a
20
1 10
Trials
1
Fig. 5. Summary of desensitization (vertical axis) and tachyphylaxis (repeat trials) of contraction induced by rapid cooling of term umbilical arteries. Cooling at first trial was from 37· C; all other trials from 25° C.
Comment Among the agonists studied, it is clear that serotonin possessed the pharmacodynamic properties most likely to be involved in closure of the extracorporeal circulation at birth. At a concentration of lO-7 mollL it produced consistently contractile responses that exceeded the maximum response to potassium chloride, indicating marked potency and efficacy. The desensitization induced by serotonin was clearly less than that of any other agonist and tachyphylaxis to serotonin was not evident. Moreover, the responses to high concentrations of serotonin lasted for hours without significant decrement and the contraction was difficult to terminate with washout, suggesting a high affinity for the receptors of the arterial wall. The mast cells of the cord, which are abundantly associated with vessels, may be one source of serotonin for closure. It has been estimated that the concentration of an agonist in the area of effector cells may reach lO-4 mol/L on the release of vesicles containing transmitters. If so, the concentration of serotonin released could be high and, in the absence of a notable circulation in the cord, remain high. Mast cells also release histamine, heparin, platelet-activating factor, proteases, chemotactic factors, and a variety of eicosanoids (prostaglandins D2 , E., D, and F2a and leukotriene C4 ), among other agents. The potency, efficacy, and lack of tachyphylaxis obtained with histamine suggest that it may also contribute significantly to closure, although the desensitization induced would appear to limit its effect
Volume 160 Number 1
to the earlier phase of the phenomenon. The results obtained with histamine and especially with serotonin suggest that the identification of their source at birth, such as platelets or components of Wharton's jelly, would provide a better understanding of the closure phenomenon. Although the only effect obtained by increasing oxygenation was vasoconstriction, the importance of oxygen to closure is equivocal. The response to a wide range of oxygen concentrations was relatively weak, was short-lived (about 7 minutes), and was not evident in most arteries (Table II). The highest incidence of response (60%) was obtained by changing the oxygen concentration from 8% to 95%, and this result agrees with a previous report. 9 Lewis lo reported that only 3 of 11 perfused human umbilical arteries constricted with elevated oxygen levels. Eltherington et al. B performed similar experiments and remarked that the peak contraction produced by oxygen was not sustained. Others have observed that increasing the oxygen supply to the umbilical arteries incrementally from 30 to 80 mm Hg or from 5% to 95% had little or no effect on the basal tone of the vessel. l!. 12 In contrast, constriction of the ductus arteriosus of the guinea pig, sheep, and cow to oxygen is a consistent finding. I•. 15 Nevertheless, in the cat the ductus arteriosus manifests tachyphylaxis to oxygen and in the dog there is no response. 11 Thus responsiveness to oxygen varies with the species and the origin of the vessel. Our findings support the clinical arguments of others that oxygen is not the singular stimulus for closure of the umbilical circulation because pulsations of the cord may cease before birth and before respiration and because the extracorporeal circulation will continue in animals whose respirations commence before placental detachment. lb Previous investigators 3 5. 10 have shown that cooling the umbilical artery, whether perfused or studied in a tissue bath, produces vasoconstriction and that this response occurs even in arteries that are refractory to oxygen. Although this response might contribute to closure in cool climates, the present study shows that desensitization would limit the effect of cooling to only several minutes. The response was also subject to tachyphylaxis and the cooled vessel was not more sensitive to serotonin. Since the disclosure by Karim l7 that prostaglandins are formed by umbilical vessels and are vasoactive, other investigators have implicated metabolites of arachidonic acid as controlling factors of the placental circulation. Indeed, nearly all sodium arachidonate metabolites tested (prostaglandins AI, D 2, E2, Flo, F2o ), including the endoperoxides (prostaglandins G 2 and H 2 ), constrict umbilical arteries or reduce umbilical cir-
Umbilical arteries
235
Table II. Responses of term umbilical arteries to increases in oxygenation Change
In
oxygen (%)
o to 95 2 8 2 8
95 to 95 to 12 to 12 to
No. per total respondmg
4 of 2 of 6 of 1 of o of
13 10 10
4 15
Mean contraction was 0.82 ± 0.11 gm (± SEM) and percent potassium chloride was 22.8 ± 3.1 (± SEM) (n = 13). Mean duration was 6.5 ± 0.95 minutes (± SEM); geometric mean was 5.55 minutes.
culation6 . 7 (Fig. 2). The exceptions are prostaglandins E, and I., which produce diphasic effects; that is, low concentrations dilate and high ones constrict. The pharmacodynamics of the umbilical artery change as its branches penetrate the placenta and the change might be important to the closure phenomenon. The arterioles supplying term placental villi, for instance, constrict only to prostaglandin E" constrict in the presence of sodium arachidonate, and respond best to angiotensin 11.7 Since angiotensin II and sodium arachid onate were found to be more effective in the preterm umbilical artery (Fig. 2, Table I), it is possible that the term arterioles retain some of the responses evident only in the immature parent vessel. Why sodium arachid onate failed to elicit responses in the mature artery is problematic, but there are several cellular pools of esterified sodium arachidonate and one incorporates exogenous sodium arachidonate slowly. This one may predominate in the mature vessel. In any case prostaglandins are effective constrictor agents for both the umbilical artery and the villous arterioles. The endogenous production of eicosanoids by umbilical vessels does appear to represent an intrinsic control of vasomotion. This is most evident in the ductus arteriosus, which dilates to minute concentrations of prostaglandin E2 and closes in the presence of cyclooxygenase inhibitors. However, the role metabolites of sodium arachidonate play in closure shall remain unclear until the source of these eicosanoids during closure is identified. Previous reports indicate that the amount of prostaglandins formed by several milligrams of artery in 15 minutes or the amount that is present in the circulation would be insufficient to elicit the contractions we observed in vitro. 6 Moreover, the stimuli requisite for eicosanoid synthesis have not been identified. Angiotensin II, for instance, is a potent constrictor of the villous arterioles and stimulates prostaglandin synthesis, but its effect is short-lived. 7 Also, the most effective constrictor of umbilical arteries, serotonin
236 White
creatine sulfate, is unaffected by the presence of inhibitors of synthesis. 6 An unexplored possibility is that during detachment of the placenta, anoxia and proteinases, both stimuli for eicosanoid synthesis, yield a variety of vasoactive substances that constrict the resistance vessels of the villi. In this model placental vessels would be responsible for hemostasis at birth and the umbilical vessels would contribute secondarily to closure. The report l8 that hypoxia triggers constriction of the cotyledon vessels supports this posit. The absence of a contractile response to a serine protease (thrombin), to a substrate oflipoxygenase (linoleic acid), to a polypeptide (neuropeptide V), and to a component of cells and platelets (uridine triphosphate) indicates that such agents would be ineffectual in closure (Fig. 2). The relatively weak or ineffectual responses elicited by vasopressin, oxytocin, angiotensin II, acetylcholine, and catecholamines have been observed by others. 3-5 • 8 These agonists are therefore poor candidates to mediate closure. Moreover, the rapid decay of the responses produced indicates that these agonists could contribute only to the earliest phase of closure (Table I). In this regard bradykinin was among the most effective of the agonists studied, but the accelerated decay in the response after a second application indicates it also would not maintain constriction for prolonged periods. Although the most important clinical feature of closure is that extracorporeal flow ceases 3 minutes after birth, the fact that contracture of the cord vessels is reversible 3 indicates that closure is a dynamic function of vascular smooth muscle. Because this reversal is achieved in perfused cords, it is possible that some unidentified agent in cord blood maintains closure. Thrombin is apparently not one of these, but it is interesting that solutions of hemoglobin produce constriction of perfused umbilical arteries. 3 Hillier and Karim l6 reported that preterm (13 to 24 weeks) umbilical arteries failed to respond to prostaglandins El> F2, F la , and F2a and only a few responded to serotonin creatine sulfate. However, the highest concentration of prostaglandin tested (0.6 fLgi ml) was about 2 X lO-6 mollL, which the present study indicates is near threshold, and the vessels used herein were older (23 to 35 weeks). It also is not known whether the arteries they obtained from legally aborted fetuses responded to potassium chloride. In any case our findings agree that the responsiveness of the mature vessel to prostaglandins is greater than that of the preterm artery. Moreover, the consistent responses of intraabdominal umbilical veins that are 21 to 23 weeks old to serotonin creatine sulfate and acetylcholine observed by Ehinger et al. I9 support our findings that the contractile mechanisms of umbilical vessels are operative early in gestation.
January 1989 Am J Obstet Gynecol
The comparative study further indicates that the receptors or transductional mechanisms of vascular smooth muscle for contractile agonists can mature at different times during gestation and even wane with development. The contractile response to norepinephrine, arginine vasopressin, and prostaglandins E2 and D2 was clearly most evident in the term artery, but the best responses to angiotensin II and sodium arachidonate were confined to the preterm vessel (Fig. 2). On the other hand, similar responses were obtained in both sets of vessels with bradykinin, oxytocin, histamine, serotonin creatine sulfate, acetylcholine, and prostaglandin F2a • The greater force of contraction to potassium chloride in the mature vessel was expected and is probably related to muscle mass. Although the mechanism responsible for closure has not been identified, serotonin may play a pivotal role because it best exhibits the requisite pharmacodynamic properties of potency, efficacy, and duration of action. Elements of Wharton's jelly and blood could be sources of this and other significant spasmogens. Some of these may act synergistically to effect closure, but further investigation is required to elucidate events that may trigger the release of the vasoactive agents. I appreciate the advice of Henrietta S. Bada, MD, and the expertise of Mrs. Gwendolyn Stornes and Mrs. Marion Johnson during this study. REFERENCES
I. Moinian M, Meyer WW, Lind J. Diameters of umbilical cord vessels and the weight of the cord in relation to damping time. AM J OBSTET GYNECOL 1969; 105:604. 2. Stembera ZK, Hodr J, Janda J. Umbilical blood flow in healthy newborn infants during the first minutes after birth. AMJ OBSTET GYNECOL 1965;91:568. 3. von Euler US. Action _of adrenaline, acetylcholine and other substances on nerve-free vessels (human placenta). J Physiol 1938;93:129. 4. Somlyo AV, Woo C-Y, Somlyo AP. Responses of nervefree vessels to vasoactive amines and polypeptides. Am J Physiol 1965;208:748. 5. Gokhale SO, Gulati 00, Kelkar LV, Kelkar VV. Effect of some drugs on human umbilical artery in vitro. Br J Pharmacol 1966;27:332. 6. Tuvemo T, Strandberg K, Hamberg M, Samuelsson B. Maintenance of the tone of the human umbilical artery by prostaglandin and thromboxane formation. In: Samuelsson B, Paoletti R, eds. Advances in prostaglandins and thromboxane research. New York: Raven Press, 1976 vol 1:425.
7. Tulenko TN. The actions of prostaglandins and cydooxygenase inhibition on the resistance vessels supplying the human fetal placenta. Prostaglandins 1981 ;21: 1033. 8. Eltherington LG, Stoff J, Hughes T, Melmon KL. Constriction of human umbilical arteries: interaction between oxygen and bradykinin. Circ Res 1968;22:747. 9. Bor I, Guntheroth WG. In vitro response to oxygen of human umbilical arteries and of animal ductus arteriosus. Can J Physiol Pharmacol 1970;48:500. 10. Lewis BV. The response of isolated sheep and human umbilical arteries to oxygen and drugs. J Obstet Gynaecol Br Commonw 1968;75:87. II. Bj¢ro K, Haugen G, Stray-Pedersen S. Altered prostanoid
Umbilical arteries
Volume I(}() I'\\lmber I
12.
13. 14 15. Ifi.
formation m human umbilical vasculature m re~ponse to variations in oxygen tension. Prostaglandins 1987;34:377. Starling MB. Elliott RB. The effects of prostaglandins. prmtaglandin inhibitors. and oxygen on the closure of the ductus arteriosus. pulmonary arteries. and umbilical arteries 11/ <'Itl() Prostaglandins 1974;8: 187. Fa\ FS. Guinea pig ductus arteriosus. J. Cellular and metabolic basis for oxygen sensitiVity. Am .J Physiol 1971: 221 :470. (;illman RG. Burton AC. Constriction of neonatal aorta b\ raised oxygen temion. Circ Res 1966:14:755. Kovalcik V. The response of the isolated ductus arteriosus to oxygen dnd anoxia. J Physio1l963: 169: IR5. Hillier K. Karim SMM. Effects of prostaglandms E,. E,.
Fla , F' a on isolated human umbilical and placental blood vessels . .J Obstet Gynae,col Br Commonw 1968;75:667. 17. Karim SMM. The identification of prostaglandins in human umbilical cord. Br J Pharmacol 1967;29:230. 18. Howard RB, Hosokawa T, Maguire MH. Hypoxiainduced fetoplacental vasoconstriction in perfused human placental cotyledons. AM.J OBSTET GYNECOL 1987; 157: 1261. 19. Ehinger B, Gennser G, Ow man C, Persson H, Sjoberg N-O. Histochemical and pharmacological studies on amine mechanisms in the umbilical cord. umbilical vein and ductus venosus of the human fetus. Acta Physiol Scand 1968;72:15.
The effect of estrogen on placental delivery after fetectomy in baboons Eugene D. Albrecht, PhD; M. Carlyle Crenshaw, Jr., MD; and Gerald J. Pepe, PhD b Baltimore. Maryland, and Norfolk. Virginia In baboons, the placenta remains in situ and functional with respect to the potential for aromatization after removal of the fetus (fetectomy). Fetectomy therefore was used to study effects of the fetus and estrogen on placental delivery. By term, serum estradiol levels in untreated, intact baboons had increased to 4 to 8 ng/ml, and fetoplacental delivery occurred on day 184 ± 1 (mean ± SE). Fetectomy at midgestation resulted in a nondetectable serum estradiol level and a marked decline in progesterone level; however, placentas were maintained in situ and were delivered on day 171 ± 6. After fetectomy therefore the initiation of placental delivery and, presumably, myometrial contractility did not require an elevation in estrogen. Administration of estradiol (1 to 10 mg/day) to baboons after fetectomy resulted in normal serum estradiol concentrations, but placental delivery was prevented. When estrogen was discontinued on days 215 to 250, the serum estradiol level declined, and placental delivery occurred on day 262 ± 18, a value greater than in intact baboons or untreated baboons after fetectomy (p < 0.001). Thus estrogen prevented placental delivery in baboons after fetectomy. (AM J OBSTET GVNECOL 1989;160:237-41.)
Key words: Parturition, delivery, estrogen, fetus, baboon Although the presence of the fetus and increased estrogen levels may be important to the initiation of parturition in several nonprimate species,' 2 their roles have been less clearly defined in primate pregnancy. In women bearing anencephalic fetuses, the mean length of gestation did not differ significantly from the normal gestational period," although the incidence of premature and postmature births increased. Moreover
From the Department of Obstetncs and Gynecology, UmVersl(V of Maryland, School of Medlflne," and the Department of PhYSIOlogy, Eastern Vlrgmla Medical School'. RecelVed for publicatIOn April 6, 1987; reVISed Ma.v 2. 1988; acceptedJune 10.1988. Reprmt requests: Dr. Eugene D. Albraht, Department of ObstetTlcs and G,wlec%gy, UnlVenltv of Maryland School oj Medlcme. Brm/a Research Laboralon 11-019, 655 W Baltunore SI., BaltlY/lOre, MD 21201. .
in most cases labor did occur in rhesus monkeys after fetal adrenalectomy' or experimental fetal anencephaly,' although the precise timing of parturition was lost. Finally, when the fetuses of rhesus monkeys" 7 and baboons" were removed at midgestation and the placenta left in situ (fetectomy), delivery of the placenta alone occurred relatively close to normal term, although once again the timing of delivery was abnormal. It is suggested therefore that a normal, intact fetus is not absolutely required for initiation of labor in primates but that the fetus may be important in regulating the precise timing of labor. Because circulating estrogen concentrations have been shown to increase with advancing primate gestation, and pregnancy was prolonged in rhesus monkeys by suppressing estrogen formation:' it is suggested that estrogen is important to the initiation of labor in
237
Key words: Umbilical arteries, pharmacology, closure phenomenon
It is characteristic among mammals for the umbilical cord to remain attached to the placenta for prolonged periods after birth. Closure of the extracorporeal circulation is therefore essential for survival, and the mechanism responsible for closure must operate for indefinite periods. Histologic and blood flow studies of humans indicate that arterial constriction is more important than narrowing of the umbilical veins to closure, that vasoconstriction peaks at about 45 seconds, and that flow decreases most at about 120 seconds after birth.1. 2 However, flow continues for at least several minutes beyond this point, during which time the placenta reportedly begins to detach from the uterus. 2 Previous studies concerned with the placental circulation have identified a number of diverse physiologic stimuli that might contribute to, or elicit, closure of the extracorporeal circulation at birth. The stimuli studied include eicosanoids, norepinephrine, epinephrine, acetylcholine, serotonin, histamine, bradykinin, angiotensin II, oxytocin, vasopressin, hemoglobin, hyperoxia, and temperature.'" Certain prostaglandins, serotonin, histamine, and bradykinin appear to be the most effective contractile agents, although bradykinin From the Department of Pharmacology and the Newborn Center, Untversity of Tennessee Medical Center. Supported by National Institutes of Health, United States Public Health Service, grant NS 21405. Received for publicatIOn January 8, 1988; revised May 5. 1988; accepted June 13, 1988. Repnnt requests: RIchard P. White, PhD, Department of Phannacology, UniverSIty of Tennessee Medical Center, Memphis, TN 38163.
may induce tachyphylaxis. s However, the identification of factors responsible for closure has been largely based on acute responses of umbilical vessels to stimuli, the responses often were not quantified, and the relevancy of the duration of the response to closure was not evaluated. Since increasing oxygen tension may elicit contraction of umbilical vessels, some investigators have emphasized the role oxygen might play in closure. The results are equivocal in that the vasoconstriction produced by oxygen is reportedly inconsistent and not sustained. s, 1O Although cooling umbilical vessels appears to reliably provoke constriction," 5. 10 the duration of this response has not been studied, and mammalian births often occur in hot environs. Many vasoactive agents have been suspected to playa role in closure, but the role played by desensitization and tachyphylaxis in reducing the effectiveness of those agents has not been systematically explored. The present pharmacodynamic study was performed to characterize the contractile responses of mature umbilical arteries to a wide range of stimuli, with special attention paid to the phenomena of desensitization and tachyphylaxis. Contractile responses produced by umbilical arteries of preterm infants also were studied to ascertain whether the pharmacodynamic property of the artery changes significantly with maturation. Material and methods
Handling of the arteries. The umbilical arteries were considered mature if the infant was ;;;.38 weeks' ges-
229
230 White
January 1989 Am J Obstet Gynecol
5
-S
n =20
Term
4
*
.1----1-----1 Preterm
oc:: 3
.~
.=02 c::
8
1
*
,, ,, I
* ,,
, ,,
I
I
I
1
o~~~~--~--~
K+ 10
30
50
90 mM
Fig. 1. Dose-response curves comparing responses of preterm and term umbilical arteries to potassium chloride. Astensks indicate significant difference between corresponding points (P < 0.05).
tation. The outside diameter of 14 such arteries was 2.23 ± 0.08 mm. Premature arteries were obtained from 27 infants who were .:;;35 weeks' gestation. The average gestational length was 29.8 ± 1.1 weeks, and the infants weighed 1472 ± 142 gm (390 to 2540 gm). The outside diameter of the preterm arteries was 1.26 ± 0.16 mm. The vessels were routinely extirpated from the mid portion of the cord, except for a few experiments performed on umbilical arteries located within 5 cm of the abdominal wall. The lumen and outside of the arteries were washed with a physiologic salt solution and cleaned of superfluous material. A 4 mm segment (ring) was cut and slipped onto two parallel prongs of a tissue chamber. One prong was fixed and the other movable, being attached to a transducer for recording the tension of the arterial segment. The chamber was filled with 10 ml of the physiologic salt solution having the following composition (in millimoles per liter): sodium chloride, 118; potassium chloride, 4.7; magnesium sulfate, 1.2; sodium bicarbonate, 25; potassium monophosphate, monobasic, 1.2; calcium chloride, 2.5; glucose, 11.0; the pH was adjusted to 7.4 by adding hydrochloride. A mixture of 95% oxygen and 5% carbon dioxide was used to routinely aerate the tissue bath and a buffer reservoir. The bath temperature was 37° C. The arterial segment was allowed to incubate for 1.5 hours, during which time it is washed with physiologic salt solution about every 20 minutes. This wash is an irrigation from the bottom of the tissue chamber to an overflow so that the arterial ring is not exposed to air during washout. Arterial tension was initially set at 2 gm by means of a fine-positioning device (FTA 1011, Hewlett-Packard). Additional tension was applied as needed at about 20minute intervals in order to set a steady-state resting
tone of 0.5 to 1.0 gm. One hour later the responsiveness and the stability of the response for each artery were tested at least three times with 10, 30, 50, and 90 mmollL potassium chloride until a stable response was achieved. Two tissue baths were used so that arterial segments from the same individual would be studied in duplicate . The type of experiment performed on each ring differed so that any unknown peculiarity of an artery from one individual would not bias the overall findings. Drugs. The following drugs were obtained from Sigma Chemical Co., St. Louis: acetylcholine, angiotensin II, arginine vasopressin, bradykinin, histamine, neuropeptide Y, norepinephrine, oxytocin, phenylephrine, serotonin creatine sulfate, thrombin, uridine triphosphate, sodium arachidonate, linoleic acid, and prostaglandin D2 , E 2 , and F2u • These drugs were solubilized in distilled water, except that thrombin was dissolved in 0.9% sodium chloride, linoleic acid in water to which IN sodium hydroxide was carefully added, and neuropeptide Y in physiologic salt solution. The serotonin creatine sulfate antagonist cinanserin used in one study was obtained from E. R. Squibb and Sons, Inc., Princeton, N. J. The maximum volume of the solutions added to the bath per experiment was 0.2 m!. The various vehicles used to dissolve the drugs were inactive. Basic experimental design. The agonists were applied to the bath in increasing concentrations to give dose-response data. The concentrations applied ranged from 10- 7 to 10- 4 mollL, except for thrombin, which was applied as 0.1,1.0, and 10 NIH units per milliliter. The magnitude of the response was observed for 15 minutes, and its decay during that period was used as an index of desensitization. To determine whether the desensitization phenomenon persisted after the agonist was removed, the vessel was washed several times, and a period of 20 to 30 minutes elapsed to permit the artery to return to its original state. Then the agonist was applied again and the second response observed for 15 minutes. If the second response was statistically less than the first, the artery was considered to have manifested tachyphylaxis to the agent. Arteries that manifested complete tachyphylaxis to one agent, or failed to respond to a particular agonist, were exposed in addition to 90 mmoll L potassium chloride to ascertain whether the contractile mechanism was still operative. The maximum response to potassium chloride was used as a standard for comparing responses elicited by other agonists because it is independent of receptors and because of the great variation in contractile force generated among isolated arteries. Responses to the agonists were expressed as mean ± SEM gram force or percent of maximum response to potassium chlo-
Umbilical arteries
Volume 160 Number I
T
~ft
~
~ E
1 1 0
i9 II
.'
!Q
c
:~
T
0
J
e
50
1 1 1 1 1 1
P 1 1 1 1 •
6~
it 1
~P 1 1 1 1 1 1 0
e
6 P
,!
· ~ d T
II II II
\'
1 1 1 1
1 1 1
II
1
0
6 1 1
po
1
'~
• •o o
P - Preterm Arteries T - Term Arteries
J
1
pi
1 1 1
: P
~~ 1 1 1
6
1 1
•
10-7 M
1 1 1 1
J
":6 •
10-6M
J
,
•
Sundry
Av. of Max Response Response at 10"M 10-5M
T
P
"J 1 "J 1
o
.
T
J J
:::!1 0
~ 8
J
9' : II 1
LI>
100 1 + I> ::.::
U
Fatty Acids
Amlnes
Pepbdes
150
231
~ 1*
1 1 1 1 1 1
~~ 1 ri1T
c:
15
E
e
~
PT PT
Fig. 2. Summary of acute effects of agonists. BK, Bradykinin; ANG, angiotensin II; OXY, oxytocin; AVP, arginine vasopressin; NPY, neuropeptide Y; Hist, histamine; 5-HT, serotonin creatine sulfate; ACh, acetylcholine; NE, norepinephrine; PGF2., prostaglandin F2.; PGE" prostaglandin E2 ; PGD2 , prostaglandin D2 ; AA, arachidonic acid; LA, linoleic acid; UTP, uridine triphosphate. Asterisks indicate significant differences between average maximum response obtained in preterm and term umbilical arteries. Highest contraction and highest concentration shown may not always correspond because occasionally individual vessels responded best to a previous concentration. Except for NPY (n = 3), each bar graph represents 5 to 13 experiments. For clarity all SEM were omiu!'!d. Concentration of thrombin (0.1 to IOU / ml) is not represented in legend.
ride. The level of significance among the different responses was assessed by the appropriate Student t test.
Results Comparison of magnitude of responses by preterm and term arteries to agonists. Fig. 1 illustrates that the force (grams) of contraction elicited by potassium chloride was on average significantly greater in the term artery than in the pre term vessel. Although it is evident from the standard errors shown that the magnitude of the responses varied considerably from one individual to another, the arterial reaction to potassium chloride in each was remarkably stable and did not exhibit tachyphylaxis. Fig. 2 summarizes the responses elicited by agonists in preterm and term umbilical arteries. It is evident that the maximum contractile response elicited by bradykinin, oxytocin, histamine, serotonin, acetylcholine, and prostaglandin F2• were comparable in both groups of arteries when normalized to potassium chloride. However, arginine vasopressin, norepinephrine, prostaglandin E" and prostaglandin D2 produced significantly (p < 0.05) greater effects in the mature vessel than in the preterm one. On the other hand, angiotensin II and arachidonic acid produced significantly greater contractile responses in the preterm artery than in the mature one.
It should also be noted that at 10- 6 mol/L the preterm vessel was more reactive than the term artery to both angiotensin II and oxytocin. Increasing the concentration tenfold produced no further response to angiotensin II but increased the response to oxytocin significantly in the term artery (Fig. 2). In general, the response produced by 12 of the agonists at 10- 6 mollL was relatively weak, being only "",50% of that obtained with potassium chloride (Fig. 2). It is further evident that neuropeptide Y, uridine triphosphate, linoleic acid, and thrombin were not contractile agents in umbilical arteries. In contrast, bradykinin, histamine, and serotonin creatine sulfate produced responses that were> 50% of the potassium chloride induced response at 10- 6 mollL. Decay of response with time and tachyphylaxis. In these experiments the decrease in the maximal contractile response that occurred over a I5-minute interval was used as an index of desensitization. After washout and time allowed for tone to return to precontracted levels, the agonists were applied a second time. Any decrement in the second response to the agonist over the first was taken as an index of tachyphylaxis. Table I summarizes the findings. The responses produced by all of the peptides (Table I) decayed significantly and most induced tachyphylaxis. Moreover, the decay of the second response was
232 White
January 1989 Am J Obstet Gynecol
Table I. Comparison of maximum response, desensitization, and tachyphylaxis (recovery) elicited by agonists in preterm and term umbilical arteries* Response as % maximum response to potassium chlonde Agonist
Peptides Bradykinin Angiotensin II Oxytocin Arginine vasopressin Vasoactive amines 5-Hydroxytryptamine Histamine Acetylcholine Norepinephrine Fatty acids Prostaglandin F2a Prostaglandin
E2
Prostaglandin
D2
Arachidonic acid
Type of artery
Maximum
Preterm Term Preterm Term Preterm Term Preterm Term
113.8 107.2 31.4 1.6 28.8 23.8 23.6 49.5
Preterm Term Preterm Term Preterm Term Preterm Term
134.9 136.8 133.9 97.6 44.6 56.3
Preterm Term Preterm Term Preterm Term Preterm Term
106.6 118.8 71.2 119.1 63.4 108.4 49.9 4.3
± ± ± ± ± ± ± ±
1
At 15 min
21 25 17 0.8 9 7 6 18
28.3 42.6 7.0 0.2 5.0 4.4 6.4 4.2
± 18 ± 16 ± 13 ± 15 ± 14 ± 17 0.0 39.7 ± 16
94.5 118.0 51.7 42.9 5.8 17.5
± ± ± ± ± ± ± ±
19 22 16 21 21 17 24 2
± ± ± ± ± ± ± ±
7* 6* 3* 0.4 2* 2* 3* 0.6*
± 22* ± 13 ± 17* ± 14* ± 3* ± 9* 0.0 3.8 ± 2*
68.5 64.2 27.7 36.7 20.9 68.4 4.2 0.8
± ± ± ± ± ± ± ±
18 24 12* 9* 6* 26 3* 0.6
Second maxImum response as % potassium chloride
39.5 ± 14* 103.9 ± 22 26.4 ± 13 0.0* 3.8 ± 0.7* 0.5 ± 0.4* 11.4 ± 4* 32.3 ± 9 131.6 129.9 63.6 82.1 18.3 32.0
± 24 ± 19 ± 17* ± 10 ± 13* ± 16* 0.0 8.8 ± 5.6*
109.1 ± 17 113.6 ± 22 31.1 ± 20* 130.9 ± 14 34.1 ± 15* 112.7±19 0.8 ± 0.9* 0.0
*AII responses expressed as percent of maximum response to potassium chloride; second response obtained after washout of first and basal tone had returned to precontracted state. Asterisks indicate significant (p<0.05) decay in response after 15 minutes or significant difference in magnitude of first response over second. Each response was based on six to nine experiments. Agonists ranked by class of drug and by efficacy in preterm arteries.
more accelerated than the first, suggesting that tachyphylaxis is an extension of the desenitization phenomenon (data not shown). Fig. 3, B, illustrates the marked desensitization and tachyphylaxis produced by bradykinin in one preterm artery. All vasoactive amines (Table I) except serotonin creatine sulfate significantly induced desensitization. In addition, serotonin creatine sulfate was the only agonist whose effect persisted after washout, a phenomenon illustrated in Fig. 3, C. The persistent contraction after washout was evidently caused by serotonin creatine sulfate because it was ultimately removed by frequent washings and because it could be completely reversed by 10- 5 mollL cinanserin. The marked desensitization and tachyphylaxis produced by norepinephrine in the term artery was replicated with phenylephrine. Phenylephrine was studied because it reportedly produces more prolonged responses than norepinephrine in most vessels and the type of a-adrenergic receptor agonist might affect the induction of tachyphylaxis. Desensitization and tachyphylaxis to phenylephrine were nearly total (data not shown).
Among the fatty acids (Table I), prostaglandin F2a clearly produced the best response in the preterm vessel and percentage-wise induced less desensitization than the other lipids (p < 0.05 compared at 15 minutes). Desensitization to sodium arachidonate was the most rapid. Preterm arteries, but not term vessels, became tachyphylactic to prostaglandins E2 and D2 • In the term artery the prostaglandins were essentially equal in efficacy, although sodium arachidonate was ineffective. In general maturation enhanced the effectiveness of prostaglandins whereas it paradoxically reduced that of arachidonate acid. Because serotonin creatine sulfate was the only agonist not easily removed by washout and it failed to induce desensitization or tachyphylaxis, another protocol was adopted to study serotonin creatine sulfate, in which the effect of either 10- 6 (n = 6) or 10- 4 (n = 6) mollL was observed for 2.5 hours in term arteries. Contrary to expectation, desensitization to 10- 6 mollL was greater than that of 10- 4 mollL (Fig. 4). The desensitization to 10- 0 mollL was not due to a refractoriness because additional serotonin creatine sulfate pro-
Umbilical arteries 233
Volume 160 Number I
A.
Umbilical Arteries
K+
W
(:]~ • •••
B.
PT
I~
10 305090mM
5MIN
BK 1st Trial
(;] ;----• • • • -8
-7
-5(logM)
-6
2nd Trial
PT
W
~-
,. W
-8•
c.
• -6•
•
-5 (log M)
-7
5-HT
~]
• -7
~
• log M) • -5• -4(
-6
W
W
T
~ -20 MIN-
Fig_ 3. Recordings to illustrate different responses to agonists. Note sustained tonic phase of reaction to potassium chloride (A), which is readily removed with washout (W). Desensitization (BK 1st Trial) and tachyphylaxis (2nd Tnal) to contractile effects of bradykinin are clearly evident in preterm (PT) vessel. Note that sustained contraction to serotonin creatine sulfate (5-HT) was difficult to remove by washout. The 20-minute marker applies only to 5-HT.
duced a marked response (Fig. 4). Also, the more rapid decline obtained with the lower concentration was apparently not due to oxidation of serotonin creatine sulfate in the bath because lO-3 mollL ascorbic acid provided no protection against the desensitization phenomenon (n = 4). Cooling nine term umbilical arteries suddenly from 37.5" C to an average of 19.4 ± 0.6° C produced a contraction that was 48.2% ± 5.1 % of the maximum elicited by potassium chloride (5.1 ± 0.9 gm). The cooling was achieved by a wash with 8° C buffer and maintained for 10 minutes by use of a thermistor as a guide. The contraction, however, was not maintained and ceased between 5 and 10 minutes. The vessel was then warmed to 25° C and again cooled to see if less change in temperature might produce less of a response. Although the contraction was less, repeating this last procedure several times produced progressively less effect (Fig. 5). Thus cooling produced both desensitization and tachyphylaxis in these arteries (Fig. 5). Cooling the arteries slowly (15 minutes) from 37.5° to 17° C by reducing the temperature of the water jacket failed to elicit a response. Moreover, the cooled arteries were not more sensitive or responsive to serotonin creatine sulfate in concentrations of lO-8 to lO-5 mollL (n = 6).
Contractions produced by elevating oxygen content oftissue bath. In these experiments the maximum contraction to potassium chloride was established, and later
the concentration of oxygen aerating the artery was reduced from 95% to 0% (5% carbon dioxide in nitrogen), 2%, or 8% for 1.5 hour and then increased suddenly to 12% or 95%. As seen in Table II, 12 of 33 arteries responded to hyperoxia (95% oxygen), but none responded when the concentration was changed from 8% to 12%. These later concentrations provided the vessel with near physiologic levels of oxygen with bath P02 concentrations of about 63 and 95 mm Hg, respectively, and represent percentagewise the least change in oxygen tension. The contractions elicited by higher changes in oxygen content were, nevertheless, inconsistent, relatively weak, and of short duration (Table II). Incidence of responding arteries to agonists. The agonists that elicited a response in each preterm artery tested were limited to bradykinin (11 of 11), acetylcholine (8 of 8), histamine (8 of 8), serotonin creatine sulfate (13 of 13), prostaglandin E2 (9 of 9), and prostaglandin F2u (9 of9). However, the incidence of responders to angiotensin II (9 of 11), oxytocin (11 of 12), arginine vasopressin (6 of 7), sodium arachidonate (4 of 5), and prostaglandin D2 (8 of 9) was high. No response occurred with norepinephrine (0 of 7), neuropeptide Y (0 of 3), uridine triphosphate (0 of 5), or thrombin (0 of 7). The agonists that produced a response in all term vessels tested were bradykinin (35 of 35), arginine vasopressin (lO of lO), acetylcholine (lO of lO), histamine
234
White
January 1989 Am J Obstet Gynecol
i
Bath ·C 19.4
17.8
16.8
17.0
2
3
4
50
~ "#. 40
:I
:!i! 30
10"M-(6) 10" M .----.(10)
10""
aft.r
150 min
O~~--~--~--~--~--~ 5
15
30
60
150
Decay with Tlma (min)
Fig. 4. Time course of tonic phase of contraction induced by 10- 6 or 10- 4 mollL serotonin. Note that arteries exposed to 10- 6 mollL (5-HT) manifested the greatest desensitization but nevertheless responded maximally to 10- 4 mollL at 150 minutes. Asterisks represent significant differences between corresponding points (p < 0.05). Numbers in parentheses are number of experiments.
(13 of 13), norepinephrine (11 of 11), phenylephrine (12 of 12), serotonin creatine sulfate (21 of 21), prostaglandin D2 (8 of 8), prostaglandin E2 (11 of 11), and prostaglandin F2a (20 of 20). Those that produced no or a low incidence of response were angiotensin II (4 of lO), sodium arachidonate (l in 5), linoleic acid (1 of 4), thrombin (0 of 9), and uridine triphosphate (0 of 7). The capacity of angiotensin II to elicit a higher incidence of response in the preterm artery (81.8%) than in the term one (40%) was unexpected as were the effects obtained with sodium arachidonate. The fact that many agonists produced responses in all arteries tested suggests that, in the absence of further analysis, these might be considered important to the closure phenomenon. Preliminary experiments performed on umbilical arterial segments from proximal portion of cord. To ascertain whether segments of the umbilical artery near the abdominal wall might differ pharmacodynamically from those located midway in the cord, arterial rings were excised from the proximal portion of the cord, within 5 cm of the abdomen. These cords were obtained from eight preterm and four term infants. The following compounds were applied to at least two arteries from different individuals: bradykinin, oxytocin, angiotensin II, arginine vasopressin, histamine, norepinephrine, serotonin creatine sulfate, prostaglandins D2 and E2 , and thrombin. Since the responses produced did not differ in any obvious manner from those obtained from arterial segments more distal in the cord, these experiments were discontinued.
8
.s
a
20
1 10
Trials
1
Fig. 5. Summary of desensitization (vertical axis) and tachyphylaxis (repeat trials) of contraction induced by rapid cooling of term umbilical arteries. Cooling at first trial was from 37· C; all other trials from 25° C.
Comment Among the agonists studied, it is clear that serotonin possessed the pharmacodynamic properties most likely to be involved in closure of the extracorporeal circulation at birth. At a concentration of lO-7 mollL it produced consistently contractile responses that exceeded the maximum response to potassium chloride, indicating marked potency and efficacy. The desensitization induced by serotonin was clearly less than that of any other agonist and tachyphylaxis to serotonin was not evident. Moreover, the responses to high concentrations of serotonin lasted for hours without significant decrement and the contraction was difficult to terminate with washout, suggesting a high affinity for the receptors of the arterial wall. The mast cells of the cord, which are abundantly associated with vessels, may be one source of serotonin for closure. It has been estimated that the concentration of an agonist in the area of effector cells may reach lO-4 mol/L on the release of vesicles containing transmitters. If so, the concentration of serotonin released could be high and, in the absence of a notable circulation in the cord, remain high. Mast cells also release histamine, heparin, platelet-activating factor, proteases, chemotactic factors, and a variety of eicosanoids (prostaglandins D2 , E., D, and F2a and leukotriene C4 ), among other agents. The potency, efficacy, and lack of tachyphylaxis obtained with histamine suggest that it may also contribute significantly to closure, although the desensitization induced would appear to limit its effect
Volume 160 Number 1
to the earlier phase of the phenomenon. The results obtained with histamine and especially with serotonin suggest that the identification of their source at birth, such as platelets or components of Wharton's jelly, would provide a better understanding of the closure phenomenon. Although the only effect obtained by increasing oxygenation was vasoconstriction, the importance of oxygen to closure is equivocal. The response to a wide range of oxygen concentrations was relatively weak, was short-lived (about 7 minutes), and was not evident in most arteries (Table II). The highest incidence of response (60%) was obtained by changing the oxygen concentration from 8% to 95%, and this result agrees with a previous report. 9 Lewis lo reported that only 3 of 11 perfused human umbilical arteries constricted with elevated oxygen levels. Eltherington et al. B performed similar experiments and remarked that the peak contraction produced by oxygen was not sustained. Others have observed that increasing the oxygen supply to the umbilical arteries incrementally from 30 to 80 mm Hg or from 5% to 95% had little or no effect on the basal tone of the vessel. l!. 12 In contrast, constriction of the ductus arteriosus of the guinea pig, sheep, and cow to oxygen is a consistent finding. I•. 15 Nevertheless, in the cat the ductus arteriosus manifests tachyphylaxis to oxygen and in the dog there is no response. 11 Thus responsiveness to oxygen varies with the species and the origin of the vessel. Our findings support the clinical arguments of others that oxygen is not the singular stimulus for closure of the umbilical circulation because pulsations of the cord may cease before birth and before respiration and because the extracorporeal circulation will continue in animals whose respirations commence before placental detachment. lb Previous investigators 3 5. 10 have shown that cooling the umbilical artery, whether perfused or studied in a tissue bath, produces vasoconstriction and that this response occurs even in arteries that are refractory to oxygen. Although this response might contribute to closure in cool climates, the present study shows that desensitization would limit the effect of cooling to only several minutes. The response was also subject to tachyphylaxis and the cooled vessel was not more sensitive to serotonin. Since the disclosure by Karim l7 that prostaglandins are formed by umbilical vessels and are vasoactive, other investigators have implicated metabolites of arachidonic acid as controlling factors of the placental circulation. Indeed, nearly all sodium arachidonate metabolites tested (prostaglandins AI, D 2, E2, Flo, F2o ), including the endoperoxides (prostaglandins G 2 and H 2 ), constrict umbilical arteries or reduce umbilical cir-
Umbilical arteries
235
Table II. Responses of term umbilical arteries to increases in oxygenation Change
In
oxygen (%)
o to 95 2 8 2 8
95 to 95 to 12 to 12 to
No. per total respondmg
4 of 2 of 6 of 1 of o of
13 10 10
4 15
Mean contraction was 0.82 ± 0.11 gm (± SEM) and percent potassium chloride was 22.8 ± 3.1 (± SEM) (n = 13). Mean duration was 6.5 ± 0.95 minutes (± SEM); geometric mean was 5.55 minutes.
culation6 . 7 (Fig. 2). The exceptions are prostaglandins E, and I., which produce diphasic effects; that is, low concentrations dilate and high ones constrict. The pharmacodynamics of the umbilical artery change as its branches penetrate the placenta and the change might be important to the closure phenomenon. The arterioles supplying term placental villi, for instance, constrict only to prostaglandin E" constrict in the presence of sodium arachidonate, and respond best to angiotensin 11.7 Since angiotensin II and sodium arachid onate were found to be more effective in the preterm umbilical artery (Fig. 2, Table I), it is possible that the term arterioles retain some of the responses evident only in the immature parent vessel. Why sodium arachid onate failed to elicit responses in the mature artery is problematic, but there are several cellular pools of esterified sodium arachidonate and one incorporates exogenous sodium arachidonate slowly. This one may predominate in the mature vessel. In any case prostaglandins are effective constrictor agents for both the umbilical artery and the villous arterioles. The endogenous production of eicosanoids by umbilical vessels does appear to represent an intrinsic control of vasomotion. This is most evident in the ductus arteriosus, which dilates to minute concentrations of prostaglandin E2 and closes in the presence of cyclooxygenase inhibitors. However, the role metabolites of sodium arachidonate play in closure shall remain unclear until the source of these eicosanoids during closure is identified. Previous reports indicate that the amount of prostaglandins formed by several milligrams of artery in 15 minutes or the amount that is present in the circulation would be insufficient to elicit the contractions we observed in vitro. 6 Moreover, the stimuli requisite for eicosanoid synthesis have not been identified. Angiotensin II, for instance, is a potent constrictor of the villous arterioles and stimulates prostaglandin synthesis, but its effect is short-lived. 7 Also, the most effective constrictor of umbilical arteries, serotonin
236 White
creatine sulfate, is unaffected by the presence of inhibitors of synthesis. 6 An unexplored possibility is that during detachment of the placenta, anoxia and proteinases, both stimuli for eicosanoid synthesis, yield a variety of vasoactive substances that constrict the resistance vessels of the villi. In this model placental vessels would be responsible for hemostasis at birth and the umbilical vessels would contribute secondarily to closure. The report l8 that hypoxia triggers constriction of the cotyledon vessels supports this posit. The absence of a contractile response to a serine protease (thrombin), to a substrate oflipoxygenase (linoleic acid), to a polypeptide (neuropeptide V), and to a component of cells and platelets (uridine triphosphate) indicates that such agents would be ineffectual in closure (Fig. 2). The relatively weak or ineffectual responses elicited by vasopressin, oxytocin, angiotensin II, acetylcholine, and catecholamines have been observed by others. 3-5 • 8 These agonists are therefore poor candidates to mediate closure. Moreover, the rapid decay of the responses produced indicates that these agonists could contribute only to the earliest phase of closure (Table I). In this regard bradykinin was among the most effective of the agonists studied, but the accelerated decay in the response after a second application indicates it also would not maintain constriction for prolonged periods. Although the most important clinical feature of closure is that extracorporeal flow ceases 3 minutes after birth, the fact that contracture of the cord vessels is reversible 3 indicates that closure is a dynamic function of vascular smooth muscle. Because this reversal is achieved in perfused cords, it is possible that some unidentified agent in cord blood maintains closure. Thrombin is apparently not one of these, but it is interesting that solutions of hemoglobin produce constriction of perfused umbilical arteries. 3 Hillier and Karim l6 reported that preterm (13 to 24 weeks) umbilical arteries failed to respond to prostaglandins El> F2, F la , and F2a and only a few responded to serotonin creatine sulfate. However, the highest concentration of prostaglandin tested (0.6 fLgi ml) was about 2 X lO-6 mollL, which the present study indicates is near threshold, and the vessels used herein were older (23 to 35 weeks). It also is not known whether the arteries they obtained from legally aborted fetuses responded to potassium chloride. In any case our findings agree that the responsiveness of the mature vessel to prostaglandins is greater than that of the preterm artery. Moreover, the consistent responses of intraabdominal umbilical veins that are 21 to 23 weeks old to serotonin creatine sulfate and acetylcholine observed by Ehinger et al. I9 support our findings that the contractile mechanisms of umbilical vessels are operative early in gestation.
January 1989 Am J Obstet Gynecol
The comparative study further indicates that the receptors or transductional mechanisms of vascular smooth muscle for contractile agonists can mature at different times during gestation and even wane with development. The contractile response to norepinephrine, arginine vasopressin, and prostaglandins E2 and D2 was clearly most evident in the term artery, but the best responses to angiotensin II and sodium arachidonate were confined to the preterm vessel (Fig. 2). On the other hand, similar responses were obtained in both sets of vessels with bradykinin, oxytocin, histamine, serotonin creatine sulfate, acetylcholine, and prostaglandin F2a • The greater force of contraction to potassium chloride in the mature vessel was expected and is probably related to muscle mass. Although the mechanism responsible for closure has not been identified, serotonin may play a pivotal role because it best exhibits the requisite pharmacodynamic properties of potency, efficacy, and duration of action. Elements of Wharton's jelly and blood could be sources of this and other significant spasmogens. Some of these may act synergistically to effect closure, but further investigation is required to elucidate events that may trigger the release of the vasoactive agents. I appreciate the advice of Henrietta S. Bada, MD, and the expertise of Mrs. Gwendolyn Stornes and Mrs. Marion Johnson during this study. REFERENCES
I. Moinian M, Meyer WW, Lind J. Diameters of umbilical cord vessels and the weight of the cord in relation to damping time. AM J OBSTET GYNECOL 1969; 105:604. 2. Stembera ZK, Hodr J, Janda J. Umbilical blood flow in healthy newborn infants during the first minutes after birth. AMJ OBSTET GYNECOL 1965;91:568. 3. von Euler US. Action _of adrenaline, acetylcholine and other substances on nerve-free vessels (human placenta). J Physiol 1938;93:129. 4. Somlyo AV, Woo C-Y, Somlyo AP. Responses of nervefree vessels to vasoactive amines and polypeptides. Am J Physiol 1965;208:748. 5. Gokhale SO, Gulati 00, Kelkar LV, Kelkar VV. Effect of some drugs on human umbilical artery in vitro. Br J Pharmacol 1966;27:332. 6. Tuvemo T, Strandberg K, Hamberg M, Samuelsson B. Maintenance of the tone of the human umbilical artery by prostaglandin and thromboxane formation. In: Samuelsson B, Paoletti R, eds. Advances in prostaglandins and thromboxane research. New York: Raven Press, 1976 vol 1:425.
7. Tulenko TN. The actions of prostaglandins and cydooxygenase inhibition on the resistance vessels supplying the human fetal placenta. Prostaglandins 1981 ;21: 1033. 8. Eltherington LG, Stoff J, Hughes T, Melmon KL. Constriction of human umbilical arteries: interaction between oxygen and bradykinin. Circ Res 1968;22:747. 9. Bor I, Guntheroth WG. In vitro response to oxygen of human umbilical arteries and of animal ductus arteriosus. Can J Physiol Pharmacol 1970;48:500. 10. Lewis BV. The response of isolated sheep and human umbilical arteries to oxygen and drugs. J Obstet Gynaecol Br Commonw 1968;75:87. II. Bj¢ro K, Haugen G, Stray-Pedersen S. Altered prostanoid
Umbilical arteries
Volume I(}() I'\\lmber I
12.
13. 14 15. Ifi.
formation m human umbilical vasculature m re~ponse to variations in oxygen tension. Prostaglandins 1987;34:377. Starling MB. Elliott RB. The effects of prostaglandins. prmtaglandin inhibitors. and oxygen on the closure of the ductus arteriosus. pulmonary arteries. and umbilical arteries 11/ <'Itl() Prostaglandins 1974;8: 187. Fa\ FS. Guinea pig ductus arteriosus. J. Cellular and metabolic basis for oxygen sensitiVity. Am .J Physiol 1971: 221 :470. (;illman RG. Burton AC. Constriction of neonatal aorta b\ raised oxygen temion. Circ Res 1966:14:755. Kovalcik V. The response of the isolated ductus arteriosus to oxygen dnd anoxia. J Physio1l963: 169: IR5. Hillier K. Karim SMM. Effects of prostaglandms E,. E,.
Fla , F' a on isolated human umbilical and placental blood vessels . .J Obstet Gynae,col Br Commonw 1968;75:667. 17. Karim SMM. The identification of prostaglandins in human umbilical cord. Br J Pharmacol 1967;29:230. 18. Howard RB, Hosokawa T, Maguire MH. Hypoxiainduced fetoplacental vasoconstriction in perfused human placental cotyledons. AM.J OBSTET GYNECOL 1987; 157: 1261. 19. Ehinger B, Gennser G, Ow man C, Persson H, Sjoberg N-O. Histochemical and pharmacological studies on amine mechanisms in the umbilical cord. umbilical vein and ductus venosus of the human fetus. Acta Physiol Scand 1968;72:15.
The effect of estrogen on placental delivery after fetectomy in baboons Eugene D. Albrecht, PhD; M. Carlyle Crenshaw, Jr., MD; and Gerald J. Pepe, PhD b Baltimore. Maryland, and Norfolk. Virginia In baboons, the placenta remains in situ and functional with respect to the potential for aromatization after removal of the fetus (fetectomy). Fetectomy therefore was used to study effects of the fetus and estrogen on placental delivery. By term, serum estradiol levels in untreated, intact baboons had increased to 4 to 8 ng/ml, and fetoplacental delivery occurred on day 184 ± 1 (mean ± SE). Fetectomy at midgestation resulted in a nondetectable serum estradiol level and a marked decline in progesterone level; however, placentas were maintained in situ and were delivered on day 171 ± 6. After fetectomy therefore the initiation of placental delivery and, presumably, myometrial contractility did not require an elevation in estrogen. Administration of estradiol (1 to 10 mg/day) to baboons after fetectomy resulted in normal serum estradiol concentrations, but placental delivery was prevented. When estrogen was discontinued on days 215 to 250, the serum estradiol level declined, and placental delivery occurred on day 262 ± 18, a value greater than in intact baboons or untreated baboons after fetectomy (p < 0.001). Thus estrogen prevented placental delivery in baboons after fetectomy. (AM J OBSTET GVNECOL 1989;160:237-41.)
Key words: Parturition, delivery, estrogen, fetus, baboon Although the presence of the fetus and increased estrogen levels may be important to the initiation of parturition in several nonprimate species,' 2 their roles have been less clearly defined in primate pregnancy. In women bearing anencephalic fetuses, the mean length of gestation did not differ significantly from the normal gestational period," although the incidence of premature and postmature births increased. Moreover
From the Department of Obstetncs and Gynecology, UmVersl(V of Maryland, School of Medlflne," and the Department of PhYSIOlogy, Eastern Vlrgmla Medical School'. RecelVed for publicatIOn April 6, 1987; reVISed Ma.v 2. 1988; acceptedJune 10.1988. Reprmt requests: Dr. Eugene D. Albraht, Department of ObstetTlcs and G,wlec%gy, UnlVenltv of Maryland School oj Medlcme. Brm/a Research Laboralon 11-019, 655 W Baltunore SI., BaltlY/lOre, MD 21201. .
in most cases labor did occur in rhesus monkeys after fetal adrenalectomy' or experimental fetal anencephaly,' although the precise timing of parturition was lost. Finally, when the fetuses of rhesus monkeys" 7 and baboons" were removed at midgestation and the placenta left in situ (fetectomy), delivery of the placenta alone occurred relatively close to normal term, although once again the timing of delivery was abnormal. It is suggested therefore that a normal, intact fetus is not absolutely required for initiation of labor in primates but that the fetus may be important in regulating the precise timing of labor. Because circulating estrogen concentrations have been shown to increase with advancing primate gestation, and pregnancy was prolonged in rhesus monkeys by suppressing estrogen formation:' it is suggested that estrogen is important to the initiation of labor in
237