- Email: [email protected]
Response of placental vasculature to high glucose levels in the isolated human placental cotyledon
Response of placental vasculature to high glucose levels in the isolated human placental cotyledon Jonathan B. Roth, MD, James A. Thorp, MD, Sue M. Palmer, MD, Peter C. Brath, BS, Scott W. Walsh, PhD, and Sharon S. Crandell, MD Houston, Texas We report the effect of high glucose infusion on vascular resistance in isolated human placental cotyledons perfused with Krebs-Ringer-bicarbonate solution containing 80 mg/dl (4.4 mmol/L), 160 mg/dl (8.8 mmol/L), or 320 mg/dl (17.6 mmol/L) D-glucose (n = 6). Placental vascular resistance remained constant during 25-minute perfusion periods with 80 mg/dl followed by 160 mg/dl glucose solution. Subsequent perfusion with 320 mg/dl glucose produced a significant increase in placental vascular resistance. Placentas were also studied in which the placental cotyledon was sequentially perfused for 25-minute periods with solutions containing glucose at 80 mg/dl followed immediately by 320 mg/dl (n = 5). Placental vascular resistance remained constant throughout perfusion with 80 mg/dl glucose solution but increased significantly after beginning perfusion with 320 mg/dl glucose. We conclude that the increase in placental vascular resistance appears to be a function of the high glucose level rather than the duration of glucose perfusion. (AM J OSSTET GYNECOL 1990;163:1828-30.)
Key words: Placenta, glucose, vascular resistance, diabetes
Maternal hyperglycemia results in fetal hyperglycemia, fetal hyperinsulinemia, and an increased flux of glucose into the fetal circulation. These metabolic alterations appear to have a deleterious effect on fetal well-being. Phillips et aLI showed that fetal hyperglycemia produced by glucose infusion into the ovine fetus resulted in increased oxygen consumption, hypoxemia, acidemia, and fetal death in fetuses that became severely acidemic. In studying the effects of maternal hyperglycemia on the fetus, we previously showed that the hyperglycemic fetus became hypoxemic and a mixed metabolic respiratory acidemia developed. 2 In addition, umbilical blood flow was significantly decreased at maternal glucose concentrations that were two to three times the normal range for pregnant ewes. The cause of this observed decrease in umbilical blood flow during maternal-fetal hyperglycemia was unclear. Fetal hypoxemia is a potent stimulus for the release of vasoactive substances that may contribute to reduced placental blood flow. Rankin and Phernetton" reFrom the Perinatal Physiology Research Unit, Departments of Pediatrics, Obstetrics, Gynecology, and Reproduction Sciences, and Physiology and Cell Biology, University of Texas Medical School at Houston. Supported in part by Grant No. HD 20973 from the National Institute of Child Health and Human Development. Presented at the Thirty-sixth Annual Meeting of the Society for Gynecologic Investigation, San Diego, California, March 15-18,1989. Reprint requests: Jonathan B. Roth, MD, Presbyte1ian Hospital, P.O. Box 33549, 200 Hawthm'ne Lane, Charlotte, NC 28233. 6/6/24748
1828
ported that norepinephrine and prostaglandin E2 caused vasoconstriction in the placental vascular bed and thereby decreased umbilical blood flow.' Angiotensin II increases placental vascular resistance.' It is not clear by what mechanism the decrease in umbilical blood flow is initiated or sustained during maternalfetal hyperglycemia, or what if any role fetal hypoxemia plays in the redistribution of fetal cardiac output and whether this redistribution is mediated by the release of vasoactive agents. Despite extensive in vivo work in glucose metabolism in chronically catheterized ovine fetuses, few investigators have examined directly the effect of glucose on placental vascular resistance. The following in vitro study was undertaken to examine the effects of high glucose levels on vascular resistance in the human placenta.
Material and methods The method used involved simultaneous in vitro perfusion of the placental vasculature and the intervillous space in the isolated human placental cotyledon as previously described by Schneider et aL" and Thorp et aL 7 The placentas were obtained immediately after vaginal or cesarean section delivery and placed into a bath of physiologic saline solution at room temperature. An intact lobe on the fetal side of the placenta suitable for dual perfusion was selected and the arterial and venous chorionic vessels supplying that lobe were catheterized. The placenta was then secured onto a chamber and two butterfly needles were pierced through the decidual
Response of placental vasculature to glucose
Volume 163 )\ umber 6, Part I
1829
Ci ::t
'0 E
.s ~
::J
'"'"
100 75 50
Q)
It
Glucose
80 mg %
;,...
Glucose
Glucose
160 mg %
320 mg %
~
~____~____~.____~____' __~~
~
ill
$
_1_ __
r
25 0
25
50
Time (minutes) Fig. 1. Representative tracing of change in pressure of chorionic plate artery during perfusion of isolated human placental cotyledon with sequentially increasing concentrations of glucose (80 mg/dl (4.4 mmol/L), 160 mg/dl (8,8 mmollL), and 320 mg/dl (17.6 mmoIlL),
plate of the maternal side of the placenta. The fetal flow was set at 2 to 5 mil min and the maternal flow at 7 to 12 mllmin. The maternal and fetal sides of the placenta were perfused with a Krebs-Ringerphosphate-bicarbonate solution gased with 95% air and 5% carbon dioxide at 37 0 C. Pressure changes in the chorionic plate placental arteries and veins were monitored with a physiologic recorder (model 7754B, Hewlett Packard, Waltham, Mass.). Changes in resistance were calculated by dividing the difference between arterial and venous pressures by the measured flow rates. We studied six placentas (group A). The placental cotyledons were sequentially perfused for 25-minute periods with solutions that contained glucose of 80 mg/dl (4.4 mmol/L), 160 mg/dl (8.8 mmoIlL), and 320 mg/dl (17.6 mmol/L). In addition, five placentas were studied in which the placental cotyledons were sequentially perfused for 25-minute periods with solutions that contained glucose at 80 mg/dl (4.4 mmollL) followed by 320 mg/dl (17.6 mmollL) (group B). Solutions of mannitol of 80 mg/dl (4.4 mmollL), 160 mg/dl (8.8 mmoIlL), and 320 mg/dl (17.6 mmollL) were used to perfuse two additional placentas. Data were analyzed as appropriate by the paired Student t test or analysis of variance with random block due to variability between placentas. Tissue viability was determined in six placentas by measuring lactate production after 15 minutes, 60 minutes, and 120 minutes of perfusion.
Results There was no significant change in lactate production throughout the experiment, indicating that the tissue remained viable during experimental procedures. Mean lactate production at 15 minutes was 0.79, at 60 minutes 0.89, and at 120 minutes 0.97 mmollmin/gm, These values are within the range re-
Table I. Placental vascular resistance in response to perfusion with different concentrations of glucose studied in the isolated human placental cotyledon Resistance (mm Hg . minlml) Glucose (mgldl) 80 160 320
Group A mean ± SE (n
= 6)
14.17 ± 4.26 14.26 ± 4.22 30.00 ± 11.4
G1'OUp B mean ± SE (n
= 5)
16.82 ± 3.56 31.49 ± 5.25
ported by others. s Placental vascular resistance remained constant throughout a 25-minute perfusion with 80 mg/dl glucose solution and throughout a 25minute perfusion with 160 mg/ dl glucose solution. Within 12 ± 3 minutes of beginning the perfusion with 320 mg/dl glucose, a significant increase of 110% (P < 0.05) in placental vascular resistance was observed (Fig. I, Table I). In an additional five placentas studied, vascular resistance remained constant throughout a 25-minute perfusion with 80 mg/dl glucose solution. Within 11 ± 3 minutes of beginning the perfusion with 320 mg/ dl glucose, a significant increase of 87% in placental vascular resistance was observed (p < 0.01) (Fig. 2, Table I). In the two placentas perfused with mannitol, vascular resistance remained constant throughout sequential perfusions with mannitol solutions of 80 mg/dl, 160 mg/dl, and 320 mg/dl, which shows that the vasoconstriction observed with 320 mg/dl glucose was not simply an osmotic effect.
Comment Placental vascular resistance increased in response to high glucose levels. The absence of a response in the placentas perfused with mannitol indicates that this re-
1830
Roth et al.
December 1990 Am J Obstet Gynecol
OJ :r: '0 E _E_
100 -
Glucose 80 mg %
+
Glucose 320 mg %
+
~ 50~-----------------------------------~~
;
75
a:
25
Q)
o
50
25
Time (minutes)
Fig. 2. Same as Fig. I, except the 320 mg/dl glucose perfusion immediately followed the 80 mg/dl glucose perfusion.
sponse is not an osmotic effect. In addition, the increase in placental vascular resistance appears to be a function of the high glucose level and not a result of a timerelated effect of the glucose perfusion. Because there are differences in the anatomic relationship of the perfusing vessels in the sheep versus the human placenta, one cannot be certain that high glucose levels would produce similar vasoconstriction in the sheep placenta. The data do, however, suggest that the decrease in umbilical blood flow observed by Crandell et al! in the intact pregnant sheep may have resulted from changes in resistance in the placental vascular bed as a consequence of maternal hyperglycemia. The mechanism for the increase in placental vascular resistance after perfusion with high glucose levels is not known. In vivo, fetal hyperglycemia causes hypoxemia. Fetal hypoxemia is a potent stimulus for the release of vasoactive substances that may contribute to increased placental vascular resistance. Fetal hypoxemia evokes the release of vasoconstrictors such as vasopressin, renin, epinephrine, norepinephrine, and prostaglandin E2 and F2• into the circulation."·4.;.g.lo However, this cannot explain vasoconstriction in response to high glucose levels in vitro. High glucose levels caused a decrease in prostacyclin production by cultured trophoblast cells obtained from early pregnancy,11 and there is a significant decrease in prostacyclin production with no significant change in thromboxane production by placentas obtained from women with diabetes as compared with nondiabetic women. 12 Therefore, vasoconstriction in response to high glucose levels may be related to decreased prostacyclin production leading to an imbalance between thromboxane and prostacydin levels that favors the vasoconstrictive effects of thromboxane. We speculate that the increase in placental vascular resistance observed in vitro may explain in vivo the decrease in umbilical blood flow during maternal-fetal hyperglycemia observed in our laboratory in the sheep
and by Trudinger et al. 13 in the human uncontrolled diabetic patient. REFERENCES I. Phillips AF, Porte PJ, Stabiusky S, Rosenkrantz TS, Raye
2. 3. 4. 5. 6. 7.
8. 9. 10. 11.
12.
13.
JR. Effects of chronic fetal hyperglycemia upon oxygen consumption in the ovine uterus and conceptus. ] Clin Invest 1984;74:279-86. Crandell SS, Fisher D], Morriss FH. Effects of ovine maternal hyperglycemia on fetal regional blood flows and metabolism. Am] Physiol 1985;249:E454-60. Rankin ]GH, Phernetton TM. Effect of norepinephrine on the ovine umbilical circulation. Proc Soc Exp Bioi Med 1976;152:312-7. Rankin JHG. Phernetton TM. Circulatory responses of the near-term sheep fetus to prostaglandin E2 • Am] PhysioI1976;D231:760-5. Iwamoto HS, Rudolph AM, Keol LC, Heymann MA. Hemodynamic response of the sheep fetus to vasopressin infusion. Circ Res 1976;44:430-6. Schneider H, Panigel M, Dancis ]. Transfer across the perfused human placenta of antipyrine, sodium, and leucine. AM] OBSTET GYNECOL 1972; 114:822-8. Thorp ]A, Walsh SW, Brath Pc. Comparison of the vasoactive effects of leukotrienes with thromboxane mimic in the perfused human placenta. AM J OBSTET GYNECOL 1988; 159: 1376-80. Bloxam DL, Bullen BE. Condition and performance of the perfused human placental cotyledon. AM J OBSTET GYNECOL 1986;155:382-8. Robiland ]E, Weitzman RE, Burmenster L, Smith FG. Development aspects of renal response to hypoxemia in the lamb fetus. Circ Res 1981;48:128-38. Ferris TF, Stein JH, Kauffman J. Uterine blood flow and uterine renin secretion.] Clin Invest 1972;51:2827-33. Rakoczi I, Tihanyi K, Gero G, Czeh I, Rozsa I, Gati 1. Release of prostacyclin (PGI 2 ) from trophoblast in tissue culture: the effect of glucose concentration. Acta Physiol Hung 1988;71 :545-9. Walsh SW, Parisi VM. The role of prostanoids and thromboxane in the regulation of placental blood flow. In: Rosenfeld C, ed. The uterine circulation. Ithaca, New York: Perinatology Press, 1989:273-98. Trudinger B], Giles WB, Cook CM, Bombardieri.1, Collius L. Fetal umbilical artery flow velocity waveforms and placental resistance: clinical significance. Br J Obstet Gynaecol 1985;92:23-30.
Key words: Placenta, glucose, vascular resistance, diabetes
Maternal hyperglycemia results in fetal hyperglycemia, fetal hyperinsulinemia, and an increased flux of glucose into the fetal circulation. These metabolic alterations appear to have a deleterious effect on fetal well-being. Phillips et aLI showed that fetal hyperglycemia produced by glucose infusion into the ovine fetus resulted in increased oxygen consumption, hypoxemia, acidemia, and fetal death in fetuses that became severely acidemic. In studying the effects of maternal hyperglycemia on the fetus, we previously showed that the hyperglycemic fetus became hypoxemic and a mixed metabolic respiratory acidemia developed. 2 In addition, umbilical blood flow was significantly decreased at maternal glucose concentrations that were two to three times the normal range for pregnant ewes. The cause of this observed decrease in umbilical blood flow during maternal-fetal hyperglycemia was unclear. Fetal hypoxemia is a potent stimulus for the release of vasoactive substances that may contribute to reduced placental blood flow. Rankin and Phernetton" reFrom the Perinatal Physiology Research Unit, Departments of Pediatrics, Obstetrics, Gynecology, and Reproduction Sciences, and Physiology and Cell Biology, University of Texas Medical School at Houston. Supported in part by Grant No. HD 20973 from the National Institute of Child Health and Human Development. Presented at the Thirty-sixth Annual Meeting of the Society for Gynecologic Investigation, San Diego, California, March 15-18,1989. Reprint requests: Jonathan B. Roth, MD, Presbyte1ian Hospital, P.O. Box 33549, 200 Hawthm'ne Lane, Charlotte, NC 28233. 6/6/24748
1828
ported that norepinephrine and prostaglandin E2 caused vasoconstriction in the placental vascular bed and thereby decreased umbilical blood flow.' Angiotensin II increases placental vascular resistance.' It is not clear by what mechanism the decrease in umbilical blood flow is initiated or sustained during maternalfetal hyperglycemia, or what if any role fetal hypoxemia plays in the redistribution of fetal cardiac output and whether this redistribution is mediated by the release of vasoactive agents. Despite extensive in vivo work in glucose metabolism in chronically catheterized ovine fetuses, few investigators have examined directly the effect of glucose on placental vascular resistance. The following in vitro study was undertaken to examine the effects of high glucose levels on vascular resistance in the human placenta.
Material and methods The method used involved simultaneous in vitro perfusion of the placental vasculature and the intervillous space in the isolated human placental cotyledon as previously described by Schneider et aL" and Thorp et aL 7 The placentas were obtained immediately after vaginal or cesarean section delivery and placed into a bath of physiologic saline solution at room temperature. An intact lobe on the fetal side of the placenta suitable for dual perfusion was selected and the arterial and venous chorionic vessels supplying that lobe were catheterized. The placenta was then secured onto a chamber and two butterfly needles were pierced through the decidual
Response of placental vasculature to glucose
Volume 163 )\ umber 6, Part I
1829
Ci ::t
'0 E
.s ~
::J
'"'"
100 75 50
Q)
It
Glucose
80 mg %
;,...
Glucose
Glucose
160 mg %
320 mg %
~
~____~____~.____~____' __~~
~
ill
$
_1_ __
r
25 0
25
50
Time (minutes) Fig. 1. Representative tracing of change in pressure of chorionic plate artery during perfusion of isolated human placental cotyledon with sequentially increasing concentrations of glucose (80 mg/dl (4.4 mmol/L), 160 mg/dl (8,8 mmollL), and 320 mg/dl (17.6 mmoIlL),
plate of the maternal side of the placenta. The fetal flow was set at 2 to 5 mil min and the maternal flow at 7 to 12 mllmin. The maternal and fetal sides of the placenta were perfused with a Krebs-Ringerphosphate-bicarbonate solution gased with 95% air and 5% carbon dioxide at 37 0 C. Pressure changes in the chorionic plate placental arteries and veins were monitored with a physiologic recorder (model 7754B, Hewlett Packard, Waltham, Mass.). Changes in resistance were calculated by dividing the difference between arterial and venous pressures by the measured flow rates. We studied six placentas (group A). The placental cotyledons were sequentially perfused for 25-minute periods with solutions that contained glucose of 80 mg/dl (4.4 mmol/L), 160 mg/dl (8.8 mmoIlL), and 320 mg/dl (17.6 mmol/L). In addition, five placentas were studied in which the placental cotyledons were sequentially perfused for 25-minute periods with solutions that contained glucose at 80 mg/dl (4.4 mmollL) followed by 320 mg/dl (17.6 mmollL) (group B). Solutions of mannitol of 80 mg/dl (4.4 mmollL), 160 mg/dl (8.8 mmoIlL), and 320 mg/dl (17.6 mmollL) were used to perfuse two additional placentas. Data were analyzed as appropriate by the paired Student t test or analysis of variance with random block due to variability between placentas. Tissue viability was determined in six placentas by measuring lactate production after 15 minutes, 60 minutes, and 120 minutes of perfusion.
Results There was no significant change in lactate production throughout the experiment, indicating that the tissue remained viable during experimental procedures. Mean lactate production at 15 minutes was 0.79, at 60 minutes 0.89, and at 120 minutes 0.97 mmollmin/gm, These values are within the range re-
Table I. Placental vascular resistance in response to perfusion with different concentrations of glucose studied in the isolated human placental cotyledon Resistance (mm Hg . minlml) Glucose (mgldl) 80 160 320
Group A mean ± SE (n
= 6)
14.17 ± 4.26 14.26 ± 4.22 30.00 ± 11.4
G1'OUp B mean ± SE (n
= 5)
16.82 ± 3.56 31.49 ± 5.25
ported by others. s Placental vascular resistance remained constant throughout a 25-minute perfusion with 80 mg/dl glucose solution and throughout a 25minute perfusion with 160 mg/ dl glucose solution. Within 12 ± 3 minutes of beginning the perfusion with 320 mg/dl glucose, a significant increase of 110% (P < 0.05) in placental vascular resistance was observed (Fig. I, Table I). In an additional five placentas studied, vascular resistance remained constant throughout a 25-minute perfusion with 80 mg/dl glucose solution. Within 11 ± 3 minutes of beginning the perfusion with 320 mg/ dl glucose, a significant increase of 87% in placental vascular resistance was observed (p < 0.01) (Fig. 2, Table I). In the two placentas perfused with mannitol, vascular resistance remained constant throughout sequential perfusions with mannitol solutions of 80 mg/dl, 160 mg/dl, and 320 mg/dl, which shows that the vasoconstriction observed with 320 mg/dl glucose was not simply an osmotic effect.
Comment Placental vascular resistance increased in response to high glucose levels. The absence of a response in the placentas perfused with mannitol indicates that this re-
1830
Roth et al.
December 1990 Am J Obstet Gynecol
OJ :r: '0 E _E_
100 -
Glucose 80 mg %
+
Glucose 320 mg %
+
~ 50~-----------------------------------~~
;
75
a:
25
Q)
o
50
25
Time (minutes)
Fig. 2. Same as Fig. I, except the 320 mg/dl glucose perfusion immediately followed the 80 mg/dl glucose perfusion.
sponse is not an osmotic effect. In addition, the increase in placental vascular resistance appears to be a function of the high glucose level and not a result of a timerelated effect of the glucose perfusion. Because there are differences in the anatomic relationship of the perfusing vessels in the sheep versus the human placenta, one cannot be certain that high glucose levels would produce similar vasoconstriction in the sheep placenta. The data do, however, suggest that the decrease in umbilical blood flow observed by Crandell et al! in the intact pregnant sheep may have resulted from changes in resistance in the placental vascular bed as a consequence of maternal hyperglycemia. The mechanism for the increase in placental vascular resistance after perfusion with high glucose levels is not known. In vivo, fetal hyperglycemia causes hypoxemia. Fetal hypoxemia is a potent stimulus for the release of vasoactive substances that may contribute to increased placental vascular resistance. Fetal hypoxemia evokes the release of vasoconstrictors such as vasopressin, renin, epinephrine, norepinephrine, and prostaglandin E2 and F2• into the circulation."·4.;.g.lo However, this cannot explain vasoconstriction in response to high glucose levels in vitro. High glucose levels caused a decrease in prostacyclin production by cultured trophoblast cells obtained from early pregnancy,11 and there is a significant decrease in prostacyclin production with no significant change in thromboxane production by placentas obtained from women with diabetes as compared with nondiabetic women. 12 Therefore, vasoconstriction in response to high glucose levels may be related to decreased prostacyclin production leading to an imbalance between thromboxane and prostacydin levels that favors the vasoconstrictive effects of thromboxane. We speculate that the increase in placental vascular resistance observed in vitro may explain in vivo the decrease in umbilical blood flow during maternal-fetal hyperglycemia observed in our laboratory in the sheep
and by Trudinger et al. 13 in the human uncontrolled diabetic patient. REFERENCES I. Phillips AF, Porte PJ, Stabiusky S, Rosenkrantz TS, Raye
2. 3. 4. 5. 6. 7.
8. 9. 10. 11.
12.
13.
JR. Effects of chronic fetal hyperglycemia upon oxygen consumption in the ovine uterus and conceptus. ] Clin Invest 1984;74:279-86. Crandell SS, Fisher D], Morriss FH. Effects of ovine maternal hyperglycemia on fetal regional blood flows and metabolism. Am] Physiol 1985;249:E454-60. Rankin ]GH, Phernetton TM. Effect of norepinephrine on the ovine umbilical circulation. Proc Soc Exp Bioi Med 1976;152:312-7. Rankin JHG. Phernetton TM. Circulatory responses of the near-term sheep fetus to prostaglandin E2 • Am] PhysioI1976;D231:760-5. Iwamoto HS, Rudolph AM, Keol LC, Heymann MA. Hemodynamic response of the sheep fetus to vasopressin infusion. Circ Res 1976;44:430-6. Schneider H, Panigel M, Dancis ]. Transfer across the perfused human placenta of antipyrine, sodium, and leucine. AM] OBSTET GYNECOL 1972; 114:822-8. Thorp ]A, Walsh SW, Brath Pc. Comparison of the vasoactive effects of leukotrienes with thromboxane mimic in the perfused human placenta. AM J OBSTET GYNECOL 1988; 159: 1376-80. Bloxam DL, Bullen BE. Condition and performance of the perfused human placental cotyledon. AM J OBSTET GYNECOL 1986;155:382-8. Robiland ]E, Weitzman RE, Burmenster L, Smith FG. Development aspects of renal response to hypoxemia in the lamb fetus. Circ Res 1981;48:128-38. Ferris TF, Stein JH, Kauffman J. Uterine blood flow and uterine renin secretion.] Clin Invest 1972;51:2827-33. Rakoczi I, Tihanyi K, Gero G, Czeh I, Rozsa I, Gati 1. Release of prostacyclin (PGI 2 ) from trophoblast in tissue culture: the effect of glucose concentration. Acta Physiol Hung 1988;71 :545-9. Walsh SW, Parisi VM. The role of prostanoids and thromboxane in the regulation of placental blood flow. In: Rosenfeld C, ed. The uterine circulation. Ithaca, New York: Perinatology Press, 1989:273-98. Trudinger B], Giles WB, Cook CM, Bombardieri.1, Collius L. Fetal umbilical artery flow velocity waveforms and placental resistance: clinical significance. Br J Obstet Gynaecol 1985;92:23-30.