- Email: [email protected]
In vitro perfusion studies of the human placenta
In vitro perfusion studies of the human placenta VI. Evidence against active glucose transport PHILIP A. RICE JAMES E. ROURKE ROBERT E. L. NESBITT, JR . .S)racuse. 1-,..leu' York Pievious studies in oui laboiatory using "in vitio" pertusion have established that glucose tiansport across the human placenta is a carrier-mediated process. It is not known whether these carriers require the expenditure of metabolic energy to function. In the experiments presented here we demonstrate in the perfused placenta that there is no reduction in the rate of glucose transport or its analogue 3-0-methyl-a-D-glucopyranoside (3MG) in the presence of 1Q-•M dinitrophenol (DNP), an uncoupler of oxidative phosphorylation. The presence of DNP, however, does cause an increase in the glucose utilization rate as well as increased lactic acid production. In order to test whether glucose transport depends on the functioning of a sodium pump system, the sodium in the perfusion system was replaced with choHne chloride. The final sodium content was 30 mEq/L. In the presence 1"'\f VI
"::I U
II"\1AJ IV'IW
c-nl"'i,
UTI .;JUUIII,ollll
""''"'~"'ontr""'+i"'" thor..,... u1ar- nn. ~.t"'t.l"'r,.......,.oo .in th.o r..,t.o. nf 'llA~ tr""'t'nc:'!'P'V"\I"t l"'nr•ru"''lorol"' tn tho \,IUII\,I'IIJIII.ICil"VII 1.11...,-IV YYCil.;J I I V UQVIU"Cii.-;JV I l l lolllllj;i 1~\lliJ V I Vllrt"" ILIQII,;;J..,...,I\ V\.IIII .... Lof.l.._.'-11 I.V O.IIV
control experiments run at normal sodium levels. Also "counter transport" of glucose was observed, a further indication that the glucose carrier mechanism does not require a sodium gradient in order to function. Since the transport rate of glucose or its analogue 3MG across the placenta is not reduced in the presence of 1o-•M DNP and is not reduced in the absence of a sodium gradient, it is unlikely that the mechanism of glucose transport is an active process requiring the expenditure of metabolic energy. (AM. J. 0BSTET. GYNECOL. 133:649, 1979.)
AN UNDERSTANDING of the glucose transport mechanism in the human placenta is fundamental to the understanding of fetal nutrition. The conclusion that glucose transport through the placenta is not by simple diffusion but by a more complicated system is based largely on the glucose transfer experiments in ungulates and rodents. Widdas 1 postulated the existence of "carriers" which in combination with the glucose molecule are able to transfer glucose through the placental membrane. In this type of carrier-mediated or carrierfacilitated transport system the molecules being transported react with a mobile component in the membrane to form a complex that is then transported
From the Department of Chemical Engineering and Materials Science, Syracuse University, and the Department of Obstetrus and Gynecology, Upstate Medical Center, State University of New York. Presented at the Twenty-fifth Annual Meeting of the Society for Gynecologic Investigation, Atlanta, Georgia, March 15-18, 1978. Reprint requests: Dr. Philip A. Rice, Department of Chemical Engineering and Materials Science, Syracuse University, New York, New York /3210. 0002-9378/79/060649+07$00.70/0
©
1979 The C. V. Mosby Co.
through the membrane. Carrier-mediated transport should exhibit the following characteristics: (I) rates of transfer faster than those of simple diffusion, (2) reduced rates of transfer at a high concentration, (3) reduced transfer rates in the presence of competitive isomers and analogues, (4) "counter transport," a transient "uphill" transfer (against a concentration gradient) caused by the addition of an analogue or isomer to the system, (5) a reduction in transfer rate in the presence of inhibitors which interfere with the transfer mechanism. In earlier manuscripts we have shown that glucose transport in the perfused human placenta is reduced at high glucose concentrations and that the presence of 3-0-methyl-a-D-glucopyranoside (3MG), a nonmetaboiizabie anaiogue of giucose, aiso causes a reduction in glucose transport. 2 We have also been able to induce the "uphiil" transport of glucose, under certain conditions, with the analogue 3MG. 3 These results conclusively demonstrate that glucose transport in the human placenta is a carrier-mediated process. It is possible, however, that this carrier system requires energy to function, i.e., that it is an active transport system. The purpose of this research is to determine if the
649
650 Rice, Rourke, and Nesbitt
Am.
FETAL OXYGENATOR ~
J
\larch 15, 1979 Obstet. Gvnecnl.
_.. _______ n IMANOMETER
JF=~b__'___:_'_·~d · ==...~~
lu
FETAL LOOP MANOMETER ARTIFICIAL -----!;:.~ UTERUS
FETAL RESERVOIR
I"ETAL PUMP
MANOMETER
MATERNAL PUMP
MATERNAL RESERVOIR
Fig. l. Schematic drawing of perfusion apparatus.
function of the glucose transport system in the human placenta does require energy. Two types of experiments were conducted to test for energy dependence. In the first set of experiments, dinitrophenol (DNP) was added to the perfusion system. Since Dl'\P shuts down oxidative phosphorylation, a decrease in glucose transport rate in the presence of DNP would imply that the transport system requires aerobic glycolysis to function. 4 In a number of tissues in which the transport of glucose is coupled to the transport of sodium by the sodium pump operating in the cell membrane. a marked reduction in rate of glucose transport can be demonstrated by reducing the sodium gradient across the cell membrane. 4 • -,This is the siruation, for example, in the gut and the kidney in which glucose transport into the cell is coupled with sodium transport out of the cell. 6 The second type of experiment presented tests the dependence of the placental glucose transport Inechanism on the sodium pump system. This has been done by substituting choline chloride for the sodium chloride in the perfusion fluid. A marked reduction of the sodium concentration should lead to a reduction in the glucose transport rate if the glucose transport mechanism is energy dependent. Method
The perfusion apparatus used in our experiments is designed to simulate as closely as possible in vitro fluid
flows at pressures which occur in vivo in the term placenta. The major unit of the apparatus is the artificial uterus which houses the placenta. Connected with the unit are two circulatory loops, one simulating the maternal Circulation and the other the fetal circulation. Each of these loops contains a fluid reservoir, a pump, a How meter, and an oxygenator. Associated with the apparatus are a set of manometers to measure the maternal, fetal, and amniotic fluid pressures and two constant temperature baths and an air heater to maintain the perfusion fluid reservoirs and the artificial uterus at 35° C. A schematic drawing of the essential parts is depicted in Fig. I. Ten fresh placentas from normal term pregnancies were used. Each specimen was examined for lesions before perfusion, and anv specimen with visible or gross abnormalities was discarded. The ped'usion apparatus and operaling procedure were the same as described in earlier papers. 7 • " Both maternal and fetal circulations \\·ere initially flushed and subsequentlv perfused at 35° C with a modified KrebsRinger bicarbonate (KRB) solution maintained at a pH of 7 A ::+:: 0. I. The initial concentration of glucose in each circulation was set by bringing the glucose level in each reservoir to the desired concentration. Throughout the experiment glucose was added to the maternal ( 1,200 cm 3 ) or fetal perfusate (800 cm 3) with a constant-rate infusion pump. The rate of addition, 0.54 gm/hr, was determined by the expected utilization rate corresponding to the initial concentr·ation in each
Volume 133 :-lumber 6
In vitro perfusion studies of placenta. VI
;;<0 a>
651
200~[
E. ISO
~
I
~160~
i'+-~ vi~ ---
~ 100
0
~
~
80~----~------~----~------~----~------~~
CONTROL
\
0.14
JLM
"'
2,4 DINITROPHENOL
~
I ~ I
Fig. 2. Glucose infusion to the maternal perfusate \vith 5.0 gm of 3~1G added to the maternal perfusate at t = 46 minutes and at t = 150 minutes. At t = lOS minutes, the maternal and fetal perfusates were replaced with fresh KRB solution. At t = 120 minutes, 50 mg of DNP was added to the maternal perfusate. o = Maternal reservoir concentration; 6. = fetal reservoir concentration; A. = fetal vein concentration.
circulation. 9 Both maternal and fetal flow rates, 600 and 45 cm 3imin, respectively, remained constant throughout each experiment. Samples ( 1.0 ml) for assay were taken periodically from the maternal reservoir, fetal reservoir, and fetal vein. Glucose concentrations were measured using the "Glucostat" kit (Worthington Biochemical Corporation) which is specific for glucose, and tests in our laboratory confirm that the assay for glucose was independent of the 3MG concentration. Total hexose (glucose plus 3MG) was measured by a method described by Feteris. 10 The average expected error in the measurement of the glucose concentration is 3%. 3MG concentrations were deterrr.ined by subtracting the glucose concentration from the total hexose concentration. Sodium concentrations were determined by flame photometry.
Results The results from four of the 10 different experiments are shown in Figs. 2 to 5. In all experiments the
steady-state glucose concentrations were established by infusing glucose into the maternai circulation and, after an equilibration period of about 30 minutes, 5.0 gm of 3MG were added to either the maternal or fetal circulation. In seven of the 10 experiments, after a period of 90 to 120 minutes, the contents of both reservoirs were exchanged with fresh perfusate. In four of these experiments the original perfusate was replaced with fresh KRB solution and 50 mg of DNP (0.14 mM final concentration) was added immediately to the maternal side. In three of the six experiments in which sodium dep.endence was being tested, the reservoirs were flushed and replaced with KRB in which the sodium chloride was replaced with choline chloride, reducing the sodium content of the perfusate from 140 to 30 mEq/L. (The remaining 30 mEq of sodium \Vas due to the sodium bicarbonate in the KRB.) In the remaining three experiments, the entire perfusion was with KRB containing choline chloride instead of sodium chloride.
652
Rice, Rourke, and Nesbitt
Am.
~ ~
a: z
1-
w ~
0
u
~
u
3
(!)
8~----~~----~-----+------+-----~
CONTROL
SODIUM CHLDRI~ REPLACED BY CHOLINE CHLORIDE
!5oor ~ 400
~ a:
~ 300
~
~ u
(!)
:::!: r
TiME, MiN
Fig. 3. Glucose infusion to the maternal perfusate with 5.0 gm of 3MG added to the maternal perfusate at t = 31 minutes and at t = 136 minutes. At t = 90 minutes, the maternal and fetal perfusates were replaced with fresh KRB solution in which choline chloride was substituted for sodium chloride. o = Maternal reservoir concentration; 6 = fetal reservoir concentration; A. = fetal vein concentration.
In the experiments shown in Figs. 2 and 3, glucose was infused into the maternal circulation and 3MG was added to the maternal circulation. In the experiments shown in Figs. 4 and 5, glucose was infused into the maternal circulation and 3MG was added to the fetal circulation. DNP was added after the perfusates were exchanged in the experiments of Figs. 2 and 4, while in the experiments of Figs. 3 and 5 choline chloride replaced the sodium chloride in the perfusate. In the experiments of Figs. 2 and 3 the maternal and fetal reservoir glucose concentrations began to separate immediately after the addition of the 3MG. The drop in the fetal reservoir glucose concentration was preceded by an even more pronounced drop in the fetal vein concentration. Except when DNP was present the maternal reservoir concentration rose upon the addition of 3MG. In both runs, during the control periods (no DNP or choline present) and during the periods when DNP and choline were present, the 3MG equilibrated across the placenta at essentially the same rate as
J.
~larch 1.5. 197~ Obstet. Gynecol.
the feLal vein 3MG concentration, in anticipation ol'Lhc rise in the fetal reservoir 3MG concentration. As the 3MG concentrations approached equilibrium. the glucose concentrations also tended to return to their previous ,·alues. The major notable difference between the control and the test periods in the runs of Fig. 2 was the decrease in the maternal and fetal reservoir glucose concentrations in the presence of DNP. In the runs of Figs. 4 and 5. the effect of the addition of 3MG is opposite that of the nms shown in Figs. 2 and 3, i.e., the maternal concentration decreases and the fetal reservoir and vein concentrations increase upon the addition of 3M G. with the increase in the vein concentration anticipating the fetal reservoir concentration. This continues for a considerable period even after the fetal concentration becomes higher than the maternal concentration, demonstrating glu,ose transport against a concentration gradient. On the other hand, the maternal and fetal 3MG concentrations equilibrate essentially at the same rate in the presence of DNP or with choline replac-ing sodium as they do under normal conditions. As the 3MG concentrations equilibrate, the glucose concentrations tend to return to their normal relative values. As before, in the presence of DNP, the maternal and fetal reservoir concentrations both decrease. In the other three experiments in which there was both a control and a test period. results were the same as those of the experiments shown in the figures. The addition of 3MG led to an immediate spreading or coming together of the maternal and fetal concentration curves, depending upon whether the 3MG was added to the maternal or fetal circulations. The magnitude of these concentration changes was the same for all experiments. In three additional experiments with choline, there was no mntrol period, but the addition of the 3MG elicited the same response as that observed in the control and test periods of the choline experiments shown in Figs. 3 and 5. Comment
All of the above results are consistent with carriermediated transport of glucose 1 • 2 in which the 3MG and the glucose compete for the same transport carrier. The steady-state conditions in our experiments (which are present before the addition of 3MG) exist when: (I) the rate of glucose utilization by the placenta is equal to the rate of glucose addition to the system and (2) the concentration of glucose in each circulation is not changing with time and there is no net transfer of l!lucose from one circulation tn thf' nthf'L F.Vf•n thnnuh no net transfer of glucose is occurring, glucose is still being transported into the placenta for utilization, with
<..J
---
------
-·-
----
-------------
-----~--
Volume !33 Number 6
In vitro perfusion studies of placenta. VI
653
200r-----~------~----~------~----~------~
CONTROL
0.14 p.M2-4 DINITROPHENOL
\
,\
\
'~~
"~
. .~., . ,~ _,! __. .___ ?. ;R . _a~_-_~ ,"'"C 8_...
..._ _... o1 0
i
40
80
120 TIME,MIN
. . .., __ 160
__
200
240
Fig. 4. Glucose infusion to the maternal perfusate with 5.0 gm of 3MG added to the fetal perfusate at t = 41 minutes and t = 151 minutes. At t = 100 minutes, the maternal and fetal perfusates were replaced with fresh KRB solution. At t = 110 minutes, 50.0 mg of DNP was added to the maternal perfusate. o = Maternal reservoir concentration; /:, = fetal reservoir concentration; .t.. = fetal vein concentration.
the rate of utilization just equalling the rate of glucose addition. Also molecular exchange of glucose occurs through the placenta from the maternal circulation to the fetal circulation as well as in the reverse direction, i.e., molecules of glucose passing from one circulation to the other are replaced by molecules transferring in the reverse direction. Alexander and associates'' observed this exchange across the placentas of sheep using 14 C-glucose, and Chenard and associates, 12 also using 14 C-glucose, demonstrated in monkeys that this molecular exchange can occur against a concentration gradient. Thus, in the experiment shown in Fig. 2, the 3MG, when added to the maternal circulation at the end of the control period, competes with -glucose for carrier sites. causing the maternal glucose concentration to in<.:rease because of the inhibition of molecular glucose exchange with the fetal circulation as well as by a reduction in the amount of glucose entering the placental tissue for metabolism. The fetal glucose also begins to decrease immediately because of the continuing mo-
lecular exchange of glucose from the fetal circulation to the maternal circulation. In addition to this, glucose is now being removed from the fetal circulation by the placental tissue to replace the reduced glucose transport into the tissue from the maternal side. As the 3MG equilibrates, the effect of its gradient lessens, and the maternal and fetal concentrations tend to come together again. The decrease in the average glucose concentration in the presence of DNP corresponds to an increase in the rate of glucose utilization from 0.54 gm/hr in the control period to about 1.1 gr/hr. Such an increase might be expected since, when DNP uncouples oxidative phosphorylation, the rate of anaerobic glycolysis increases to compensate, at least partially, for the energy lost by the uncoupling, leading to a decrease in the average glucose concentration. This explanation is supported by the doubling of the lactic acid production from 3 to 4 mmo!es/hr in the control period to 7 to 8 mmoles/hr in the presence of DNP in the two experiments in which lactic acid production was measured.
654
Rice, Rourke, and Nesbitt
Am.
:'>1arch !.'>, 1979 Gvnecol
J. Obstet.
~
fi
a:: z1-
ei...t ~
u w
r
0
~ (,!)
80 SODIUM REPLACED BY CH
CONTROL
z
0
I=
~
<1:
~ z 0...t w u
z
0
u
.... '-4.~
~ ,..,
00
r
40
200
Fig. 5. Glucose infusion into the maternal perfusate with 5.0 gm of 3MG added to the fetal perfusate at t 31 minutes and t = 151 minutes. At t 90 minutes, the maternal and fetal perfusates were replaced with fresh KRB solution in which choline chloride was substituted for sodium chloride. o = Maternal reservoir concentration; /::,. = fetal reservoir concentration; & = fetal vein concentration.
The explanation of the experiment of Fig. 3 is the same as for Fig. 2 except that during the test period there was no change in glucose utilization in the presence of choline. When the 3MG is added to the fetal side in the experiments of Figs. 4 and 5, it competes ·with glucose molecules for carrier sites on the fetal side of the placenta. This competition leads to an increase in the fetal glucose concentration because the molecular glucose exchange from the fetal to the maternal circulation is being inhibited by the competition from the molecules of 3MG. This increase in the fetal glucose concentration occurs as the result of the continued exchange of glucose from the maternal to the fetal circulation, i.e., glucose molecules continue to transfer from the maternal to the fetal circulation even though the reverse exchange has been inhibited. This net transfer of
glucose to the fetal circulation leads to a decrease in the maternal concentration. Since glucose molecules are entering the placental tissue at a reduced rate from the fetal side, the tissue concentration of glucose falls, with these molecules being replaced by glucose from the maternal side. This results in a further reduction in maternal glucose concentration. Thus, when the 3MG is added to the fetal circulation, the competition of glucose and 3MG for carrier sites causes a net transfer of glucose to the fetal circulation, which is reflected by a marked increase in the fetal vein glucose level, as well as the reservoir value, to concentration levels that are higher than the maternal concentration. For the period of time until the 3MG levels begin to equilibrate across the placenta, there is an "uphill" (against the concentration gradient) transport of glucose across the placenta, i.e., countertransport. It can be seen from Figs. 4
Volume leU Number 6
and 5 that this "uphill" or countertransport of glucose against a concentration gradient is not affected by the presence of a 0.14 mM concentration of DNP in the perfusate or by a low sodium ion concentration (30mEq/L). Also there is no reduction in the rate of :IMG transport across the placenta. Our previous work, together with the work of others, establishes that glucose is transferred in the human placenta by a carrier-mediated transport system. This system has been shown to be saturable at high glucose concentrations, to be inhibited by the analogue 3MG, and to exhibit countertransport. One remaining fundamental question is: Does this system require the expenditure of metabolic energy to transport glucose? If it is a system dependent on energy, does it obtain its energy by aerobic or anaerobic glycolysis? Is it dependent on the existence of a sodium pump as in the intestine? The experiments presented indicate that the glucose transport system of the human placenta does not require energy to function. The metabolic inhibitor DNP shuts down the process of oxidative phosphorylation. In the gut this leads to a decrease in glucose transfer rate since the energy for glucose transport is derived from aerobic glycolysis. 4 Although the presence of DNP in our perfusion system did cause an increase in glucose metabolism, it apparently did not inhibit the glucose carrier mechanism. This is evidenced bv the fact that countertransport still occurs in the presence of DNP and that 3MG still transfers at the same rate whether or not DNP is present.
In vitro perfusion studies of placenta. VI
655
Glucose transport in the intestine and the kidney also has been shown to be coupled with the transport of sodium out of the cells. 4 • 6 In this process, apparently one molecule of glucose is transferred across the cell membrane into the cell for every molecule of sodium pumped out. 1a Reducing the sodium gradient across the cell membrane in these tissues leads to a reduction in sodium flux and mnsequently to a reduction in glucose transfer rate into the cell. Our experiments show no such reduction in 3MG transport rate in the presence of reduced sodium. Further, the glucose transport system does not seem to function differently in any way with choline substituted for sodium. In particular, countertransport of glucose can be elicited in essentially the same way with or without the normal sodium ion gradient. If an energy-dependent system for glucose does exist, it must be able to derive its energy from anaerobic glycolysis. In addition the system would have to be independent of the sodium gradient across the cell walls. None of the above characteristics has been observed in other organ systems such as the intestine which do actively transport glucose. To the contrary, the presence of DNP and a reduction of the sodium gradient have been shown to reduce the 3MG transport in such systems. Consequently, we conclude that glucose transport in the placenta, while carrier mediated, is unlikely to be energy dependent.
REFERENCES !. Widdas, W. F.: Inability of diffusion to account for pla-
2.
3.
4. 5.
6. 7.
cental glucose transfer in sheep and consideration of the kinetics of a possible carrier transfer,]. Physiol. 118: 23, 1952. Rice, P. A., Rourke, J. E., and Nesbitt, R. E. L., Jr.: In vitro perfusion studies of the human placenta. IV. Some characteristics of the glucose transport system in the human placenta, Gynecol. Invest. 7: 213, 1976. Rice, P. A., Nesbitt, R. E. L., Jr., and Rourke,]. E.: In vitro perfusion studies of the human placenta. V. Counter transport of glucose induced bv an analog, Gv· ~ · necol. Invest. 7: 344, I976. Goldner, A. M .. Hajjar, J. ]., and Curran, P. F.: Effects of inhibitors on 3~0~methyl glucose transport in rabbit ileum, J. Membrane Bioi. 10: 267, 1972. Rinaldo,]. E., Jennings, B. L., Frizzell, R. A., and Schultz, S. G.: Effects of uniiateral sodium replacement on sugar transport across in vitro rabbit ileum, Am.]. Physiol. 228: 854, 1975. Silverman, M.: Glucose transport in the kidney, Biochem. Biophys. Acta 457: 303, 1976. Nesbitt, R. E. L., Jr., Rice, P. A., Rourke, J. E., Torresi, V. F., and Souchay. A. M.: In vitro perfusion studies of
8.
9.
10. 11.
12.
13.
the human placenta. A newly-designed apparatus for extracorporeal perfusion achieving dual closed circulation, Gynecol. Invest. 1: 185, 1970. Rourke,]. E., Gould, L. V., Rice, P. A., and Nesbitt, R. E. L., Jr.: In vitro perfusion studies of the human placenta. Angioradiographic study of the maternal circulation, Gynecol. Invest. 4: 50, 1973. Nesbitt, R. E. L., Jr., Rice, P. A., and Rourke, J. E.: In vitro perfusion studies of the human placenta. III. The relationships between glucose utilization rates and concentration, Gvnecol. Invest. 4: 243, 1973. Feteris, W. A:: A serum glucose method without protein precipitation, Am.]. Med. Techno!. 31: 17, 1965. .Alexander, D.P., .A.ndrews, R. D., Huggett, .A. S. G., Nixon, D. A., and Widdas, W. F.: The placental transfer of sugars in the sheep: studies with radioactive sugar, ]. Physioi. i29: 352, 1955. Chenard, F. P., Danesino, V., Hartman, W. F., Huggett, A. S. G., Paul, W .. and Reynolds, S. R. M.: The transmission of hexoses across the placenta in the human and rhesus monkey,]. Physiol. 132: 289, 1956. Schultz, S. G.: Sodium coupled solute transport by small intestine. A status report, Am. J. Phys. 233: E249, 1977.
"::I U
II"\1AJ IV'IW
c-nl"'i,
UTI .;JUUIII,ollll
""''"'~"'ontr""'+i"'" thor..,... u1ar- nn. ~.t"'t.l"'r,.......,.oo .in th.o r..,t.o. nf 'llA~ tr""'t'nc:'!'P'V"\I"t l"'nr•ru"''lorol"' tn tho \,IUII\,I'IIJIII.ICil"VII 1.11...,-IV YYCil.;J I I V UQVIU"Cii.-;JV I l l lolllllj;i 1~\lliJ V I Vllrt"" ILIQII,;;J..,...,I\ V\.IIII .... Lof.l.._.'-11 I.V O.IIV
control experiments run at normal sodium levels. Also "counter transport" of glucose was observed, a further indication that the glucose carrier mechanism does not require a sodium gradient in order to function. Since the transport rate of glucose or its analogue 3MG across the placenta is not reduced in the presence of 1o-•M DNP and is not reduced in the absence of a sodium gradient, it is unlikely that the mechanism of glucose transport is an active process requiring the expenditure of metabolic energy. (AM. J. 0BSTET. GYNECOL. 133:649, 1979.)
AN UNDERSTANDING of the glucose transport mechanism in the human placenta is fundamental to the understanding of fetal nutrition. The conclusion that glucose transport through the placenta is not by simple diffusion but by a more complicated system is based largely on the glucose transfer experiments in ungulates and rodents. Widdas 1 postulated the existence of "carriers" which in combination with the glucose molecule are able to transfer glucose through the placental membrane. In this type of carrier-mediated or carrierfacilitated transport system the molecules being transported react with a mobile component in the membrane to form a complex that is then transported
From the Department of Chemical Engineering and Materials Science, Syracuse University, and the Department of Obstetrus and Gynecology, Upstate Medical Center, State University of New York. Presented at the Twenty-fifth Annual Meeting of the Society for Gynecologic Investigation, Atlanta, Georgia, March 15-18, 1978. Reprint requests: Dr. Philip A. Rice, Department of Chemical Engineering and Materials Science, Syracuse University, New York, New York /3210. 0002-9378/79/060649+07$00.70/0
©
1979 The C. V. Mosby Co.
through the membrane. Carrier-mediated transport should exhibit the following characteristics: (I) rates of transfer faster than those of simple diffusion, (2) reduced rates of transfer at a high concentration, (3) reduced transfer rates in the presence of competitive isomers and analogues, (4) "counter transport," a transient "uphill" transfer (against a concentration gradient) caused by the addition of an analogue or isomer to the system, (5) a reduction in transfer rate in the presence of inhibitors which interfere with the transfer mechanism. In earlier manuscripts we have shown that glucose transport in the perfused human placenta is reduced at high glucose concentrations and that the presence of 3-0-methyl-a-D-glucopyranoside (3MG), a nonmetaboiizabie anaiogue of giucose, aiso causes a reduction in glucose transport. 2 We have also been able to induce the "uphiil" transport of glucose, under certain conditions, with the analogue 3MG. 3 These results conclusively demonstrate that glucose transport in the human placenta is a carrier-mediated process. It is possible, however, that this carrier system requires energy to function, i.e., that it is an active transport system. The purpose of this research is to determine if the
649
650 Rice, Rourke, and Nesbitt
Am.
FETAL OXYGENATOR ~
J
\larch 15, 1979 Obstet. Gvnecnl.
_.. _______ n IMANOMETER
JF=~b__'___:_'_·~d · ==...~~
lu
FETAL LOOP MANOMETER ARTIFICIAL -----!;:.~ UTERUS
FETAL RESERVOIR
I"ETAL PUMP
MANOMETER
MATERNAL PUMP
MATERNAL RESERVOIR
Fig. l. Schematic drawing of perfusion apparatus.
function of the glucose transport system in the human placenta does require energy. Two types of experiments were conducted to test for energy dependence. In the first set of experiments, dinitrophenol (DNP) was added to the perfusion system. Since Dl'\P shuts down oxidative phosphorylation, a decrease in glucose transport rate in the presence of DNP would imply that the transport system requires aerobic glycolysis to function. 4 In a number of tissues in which the transport of glucose is coupled to the transport of sodium by the sodium pump operating in the cell membrane. a marked reduction in rate of glucose transport can be demonstrated by reducing the sodium gradient across the cell membrane. 4 • -,This is the siruation, for example, in the gut and the kidney in which glucose transport into the cell is coupled with sodium transport out of the cell. 6 The second type of experiment presented tests the dependence of the placental glucose transport Inechanism on the sodium pump system. This has been done by substituting choline chloride for the sodium chloride in the perfusion fluid. A marked reduction of the sodium concentration should lead to a reduction in the glucose transport rate if the glucose transport mechanism is energy dependent. Method
The perfusion apparatus used in our experiments is designed to simulate as closely as possible in vitro fluid
flows at pressures which occur in vivo in the term placenta. The major unit of the apparatus is the artificial uterus which houses the placenta. Connected with the unit are two circulatory loops, one simulating the maternal Circulation and the other the fetal circulation. Each of these loops contains a fluid reservoir, a pump, a How meter, and an oxygenator. Associated with the apparatus are a set of manometers to measure the maternal, fetal, and amniotic fluid pressures and two constant temperature baths and an air heater to maintain the perfusion fluid reservoirs and the artificial uterus at 35° C. A schematic drawing of the essential parts is depicted in Fig. I. Ten fresh placentas from normal term pregnancies were used. Each specimen was examined for lesions before perfusion, and anv specimen with visible or gross abnormalities was discarded. The ped'usion apparatus and operaling procedure were the same as described in earlier papers. 7 • " Both maternal and fetal circulations \\·ere initially flushed and subsequentlv perfused at 35° C with a modified KrebsRinger bicarbonate (KRB) solution maintained at a pH of 7 A ::+:: 0. I. The initial concentration of glucose in each circulation was set by bringing the glucose level in each reservoir to the desired concentration. Throughout the experiment glucose was added to the maternal ( 1,200 cm 3 ) or fetal perfusate (800 cm 3) with a constant-rate infusion pump. The rate of addition, 0.54 gm/hr, was determined by the expected utilization rate corresponding to the initial concentr·ation in each
Volume 133 :-lumber 6
In vitro perfusion studies of placenta. VI
;;<0 a>
651
200~[
E. ISO
~
I
~160~
i'+-~ vi~ ---
~ 100
0
~
~
80~----~------~----~------~----~------~~
CONTROL
\
0.14
JLM
"'
2,4 DINITROPHENOL
~
I ~ I
Fig. 2. Glucose infusion to the maternal perfusate \vith 5.0 gm of 3~1G added to the maternal perfusate at t = 46 minutes and at t = 150 minutes. At t = lOS minutes, the maternal and fetal perfusates were replaced with fresh KRB solution. At t = 120 minutes, 50 mg of DNP was added to the maternal perfusate. o = Maternal reservoir concentration; 6. = fetal reservoir concentration; A. = fetal vein concentration.
circulation. 9 Both maternal and fetal flow rates, 600 and 45 cm 3imin, respectively, remained constant throughout each experiment. Samples ( 1.0 ml) for assay were taken periodically from the maternal reservoir, fetal reservoir, and fetal vein. Glucose concentrations were measured using the "Glucostat" kit (Worthington Biochemical Corporation) which is specific for glucose, and tests in our laboratory confirm that the assay for glucose was independent of the 3MG concentration. Total hexose (glucose plus 3MG) was measured by a method described by Feteris. 10 The average expected error in the measurement of the glucose concentration is 3%. 3MG concentrations were deterrr.ined by subtracting the glucose concentration from the total hexose concentration. Sodium concentrations were determined by flame photometry.
Results The results from four of the 10 different experiments are shown in Figs. 2 to 5. In all experiments the
steady-state glucose concentrations were established by infusing glucose into the maternai circulation and, after an equilibration period of about 30 minutes, 5.0 gm of 3MG were added to either the maternal or fetal circulation. In seven of the 10 experiments, after a period of 90 to 120 minutes, the contents of both reservoirs were exchanged with fresh perfusate. In four of these experiments the original perfusate was replaced with fresh KRB solution and 50 mg of DNP (0.14 mM final concentration) was added immediately to the maternal side. In three of the six experiments in which sodium dep.endence was being tested, the reservoirs were flushed and replaced with KRB in which the sodium chloride was replaced with choline chloride, reducing the sodium content of the perfusate from 140 to 30 mEq/L. (The remaining 30 mEq of sodium \Vas due to the sodium bicarbonate in the KRB.) In the remaining three experiments, the entire perfusion was with KRB containing choline chloride instead of sodium chloride.
652
Rice, Rourke, and Nesbitt
Am.
~ ~
a: z
1-
w ~
0
u
~
u
3
(!)
8~----~~----~-----+------+-----~
CONTROL
SODIUM CHLDRI~ REPLACED BY CHOLINE CHLORIDE
!5oor ~ 400
~ a:
~ 300
~
~ u
(!)
:::!: r
TiME, MiN
Fig. 3. Glucose infusion to the maternal perfusate with 5.0 gm of 3MG added to the maternal perfusate at t = 31 minutes and at t = 136 minutes. At t = 90 minutes, the maternal and fetal perfusates were replaced with fresh KRB solution in which choline chloride was substituted for sodium chloride. o = Maternal reservoir concentration; 6 = fetal reservoir concentration; A. = fetal vein concentration.
In the experiments shown in Figs. 2 and 3, glucose was infused into the maternal circulation and 3MG was added to the maternal circulation. In the experiments shown in Figs. 4 and 5, glucose was infused into the maternal circulation and 3MG was added to the fetal circulation. DNP was added after the perfusates were exchanged in the experiments of Figs. 2 and 4, while in the experiments of Figs. 3 and 5 choline chloride replaced the sodium chloride in the perfusate. In the experiments of Figs. 2 and 3 the maternal and fetal reservoir glucose concentrations began to separate immediately after the addition of the 3MG. The drop in the fetal reservoir glucose concentration was preceded by an even more pronounced drop in the fetal vein concentration. Except when DNP was present the maternal reservoir concentration rose upon the addition of 3MG. In both runs, during the control periods (no DNP or choline present) and during the periods when DNP and choline were present, the 3MG equilibrated across the placenta at essentially the same rate as
J.
~larch 1.5. 197~ Obstet. Gynecol.
the feLal vein 3MG concentration, in anticipation ol'Lhc rise in the fetal reservoir 3MG concentration. As the 3MG concentrations approached equilibrium. the glucose concentrations also tended to return to their previous ,·alues. The major notable difference between the control and the test periods in the runs of Fig. 2 was the decrease in the maternal and fetal reservoir glucose concentrations in the presence of DNP. In the runs of Figs. 4 and 5. the effect of the addition of 3MG is opposite that of the nms shown in Figs. 2 and 3, i.e., the maternal concentration decreases and the fetal reservoir and vein concentrations increase upon the addition of 3M G. with the increase in the vein concentration anticipating the fetal reservoir concentration. This continues for a considerable period even after the fetal concentration becomes higher than the maternal concentration, demonstrating glu,ose transport against a concentration gradient. On the other hand, the maternal and fetal 3MG concentrations equilibrate essentially at the same rate in the presence of DNP or with choline replac-ing sodium as they do under normal conditions. As the 3MG concentrations equilibrate, the glucose concentrations tend to return to their normal relative values. As before, in the presence of DNP, the maternal and fetal reservoir concentrations both decrease. In the other three experiments in which there was both a control and a test period. results were the same as those of the experiments shown in the figures. The addition of 3MG led to an immediate spreading or coming together of the maternal and fetal concentration curves, depending upon whether the 3MG was added to the maternal or fetal circulations. The magnitude of these concentration changes was the same for all experiments. In three additional experiments with choline, there was no mntrol period, but the addition of the 3MG elicited the same response as that observed in the control and test periods of the choline experiments shown in Figs. 3 and 5. Comment
All of the above results are consistent with carriermediated transport of glucose 1 • 2 in which the 3MG and the glucose compete for the same transport carrier. The steady-state conditions in our experiments (which are present before the addition of 3MG) exist when: (I) the rate of glucose utilization by the placenta is equal to the rate of glucose addition to the system and (2) the concentration of glucose in each circulation is not changing with time and there is no net transfer of l!lucose from one circulation tn thf' nthf'L F.Vf•n thnnuh no net transfer of glucose is occurring, glucose is still being transported into the placenta for utilization, with
<..J
---
------
-·-
----
-------------
-----~--
Volume !33 Number 6
In vitro perfusion studies of placenta. VI
653
200r-----~------~----~------~----~------~
CONTROL
0.14 p.M2-4 DINITROPHENOL
\
,\
\
'~~
"~
. .~., . ,~ _,! __. .___ ?. ;R . _a~_-_~ ,"'"C 8_...
..._ _... o1 0
i
40
80
120 TIME,MIN
. . .., __ 160
__
200
240
Fig. 4. Glucose infusion to the maternal perfusate with 5.0 gm of 3MG added to the fetal perfusate at t = 41 minutes and t = 151 minutes. At t = 100 minutes, the maternal and fetal perfusates were replaced with fresh KRB solution. At t = 110 minutes, 50.0 mg of DNP was added to the maternal perfusate. o = Maternal reservoir concentration; /:, = fetal reservoir concentration; .t.. = fetal vein concentration.
the rate of utilization just equalling the rate of glucose addition. Also molecular exchange of glucose occurs through the placenta from the maternal circulation to the fetal circulation as well as in the reverse direction, i.e., molecules of glucose passing from one circulation to the other are replaced by molecules transferring in the reverse direction. Alexander and associates'' observed this exchange across the placentas of sheep using 14 C-glucose, and Chenard and associates, 12 also using 14 C-glucose, demonstrated in monkeys that this molecular exchange can occur against a concentration gradient. Thus, in the experiment shown in Fig. 2, the 3MG, when added to the maternal circulation at the end of the control period, competes with -glucose for carrier sites. causing the maternal glucose concentration to in<.:rease because of the inhibition of molecular glucose exchange with the fetal circulation as well as by a reduction in the amount of glucose entering the placental tissue for metabolism. The fetal glucose also begins to decrease immediately because of the continuing mo-
lecular exchange of glucose from the fetal circulation to the maternal circulation. In addition to this, glucose is now being removed from the fetal circulation by the placental tissue to replace the reduced glucose transport into the tissue from the maternal side. As the 3MG equilibrates, the effect of its gradient lessens, and the maternal and fetal concentrations tend to come together again. The decrease in the average glucose concentration in the presence of DNP corresponds to an increase in the rate of glucose utilization from 0.54 gm/hr in the control period to about 1.1 gr/hr. Such an increase might be expected since, when DNP uncouples oxidative phosphorylation, the rate of anaerobic glycolysis increases to compensate, at least partially, for the energy lost by the uncoupling, leading to a decrease in the average glucose concentration. This explanation is supported by the doubling of the lactic acid production from 3 to 4 mmo!es/hr in the control period to 7 to 8 mmoles/hr in the presence of DNP in the two experiments in which lactic acid production was measured.
654
Rice, Rourke, and Nesbitt
Am.
:'>1arch !.'>, 1979 Gvnecol
J. Obstet.
~
fi
a:: z1-
ei...t ~
u w
r
0
~ (,!)
80 SODIUM REPLACED BY CH
CONTROL
z
0
I=
~
<1:
~ z 0...t w u
z
0
u
.... '-4.~
~ ,..,
00
r
40
200
Fig. 5. Glucose infusion into the maternal perfusate with 5.0 gm of 3MG added to the fetal perfusate at t 31 minutes and t = 151 minutes. At t 90 minutes, the maternal and fetal perfusates were replaced with fresh KRB solution in which choline chloride was substituted for sodium chloride. o = Maternal reservoir concentration; /::,. = fetal reservoir concentration; & = fetal vein concentration.
The explanation of the experiment of Fig. 3 is the same as for Fig. 2 except that during the test period there was no change in glucose utilization in the presence of choline. When the 3MG is added to the fetal side in the experiments of Figs. 4 and 5, it competes ·with glucose molecules for carrier sites on the fetal side of the placenta. This competition leads to an increase in the fetal glucose concentration because the molecular glucose exchange from the fetal to the maternal circulation is being inhibited by the competition from the molecules of 3MG. This increase in the fetal glucose concentration occurs as the result of the continued exchange of glucose from the maternal to the fetal circulation, i.e., glucose molecules continue to transfer from the maternal to the fetal circulation even though the reverse exchange has been inhibited. This net transfer of
glucose to the fetal circulation leads to a decrease in the maternal concentration. Since glucose molecules are entering the placental tissue at a reduced rate from the fetal side, the tissue concentration of glucose falls, with these molecules being replaced by glucose from the maternal side. This results in a further reduction in maternal glucose concentration. Thus, when the 3MG is added to the fetal circulation, the competition of glucose and 3MG for carrier sites causes a net transfer of glucose to the fetal circulation, which is reflected by a marked increase in the fetal vein glucose level, as well as the reservoir value, to concentration levels that are higher than the maternal concentration. For the period of time until the 3MG levels begin to equilibrate across the placenta, there is an "uphill" (against the concentration gradient) transport of glucose across the placenta, i.e., countertransport. It can be seen from Figs. 4
Volume leU Number 6
and 5 that this "uphill" or countertransport of glucose against a concentration gradient is not affected by the presence of a 0.14 mM concentration of DNP in the perfusate or by a low sodium ion concentration (30mEq/L). Also there is no reduction in the rate of :IMG transport across the placenta. Our previous work, together with the work of others, establishes that glucose is transferred in the human placenta by a carrier-mediated transport system. This system has been shown to be saturable at high glucose concentrations, to be inhibited by the analogue 3MG, and to exhibit countertransport. One remaining fundamental question is: Does this system require the expenditure of metabolic energy to transport glucose? If it is a system dependent on energy, does it obtain its energy by aerobic or anaerobic glycolysis? Is it dependent on the existence of a sodium pump as in the intestine? The experiments presented indicate that the glucose transport system of the human placenta does not require energy to function. The metabolic inhibitor DNP shuts down the process of oxidative phosphorylation. In the gut this leads to a decrease in glucose transfer rate since the energy for glucose transport is derived from aerobic glycolysis. 4 Although the presence of DNP in our perfusion system did cause an increase in glucose metabolism, it apparently did not inhibit the glucose carrier mechanism. This is evidenced bv the fact that countertransport still occurs in the presence of DNP and that 3MG still transfers at the same rate whether or not DNP is present.
In vitro perfusion studies of placenta. VI
655
Glucose transport in the intestine and the kidney also has been shown to be coupled with the transport of sodium out of the cells. 4 • 6 In this process, apparently one molecule of glucose is transferred across the cell membrane into the cell for every molecule of sodium pumped out. 1a Reducing the sodium gradient across the cell membrane in these tissues leads to a reduction in sodium flux and mnsequently to a reduction in glucose transfer rate into the cell. Our experiments show no such reduction in 3MG transport rate in the presence of reduced sodium. Further, the glucose transport system does not seem to function differently in any way with choline substituted for sodium. In particular, countertransport of glucose can be elicited in essentially the same way with or without the normal sodium ion gradient. If an energy-dependent system for glucose does exist, it must be able to derive its energy from anaerobic glycolysis. In addition the system would have to be independent of the sodium gradient across the cell walls. None of the above characteristics has been observed in other organ systems such as the intestine which do actively transport glucose. To the contrary, the presence of DNP and a reduction of the sodium gradient have been shown to reduce the 3MG transport in such systems. Consequently, we conclude that glucose transport in the placenta, while carrier mediated, is unlikely to be energy dependent.
REFERENCES !. Widdas, W. F.: Inability of diffusion to account for pla-
2.
3.
4. 5.
6. 7.
cental glucose transfer in sheep and consideration of the kinetics of a possible carrier transfer,]. Physiol. 118: 23, 1952. Rice, P. A., Rourke, J. E., and Nesbitt, R. E. L., Jr.: In vitro perfusion studies of the human placenta. IV. Some characteristics of the glucose transport system in the human placenta, Gynecol. Invest. 7: 213, 1976. Rice, P. A., Nesbitt, R. E. L., Jr., and Rourke,]. E.: In vitro perfusion studies of the human placenta. V. Counter transport of glucose induced bv an analog, Gv· ~ · necol. Invest. 7: 344, I976. Goldner, A. M .. Hajjar, J. ]., and Curran, P. F.: Effects of inhibitors on 3~0~methyl glucose transport in rabbit ileum, J. Membrane Bioi. 10: 267, 1972. Rinaldo,]. E., Jennings, B. L., Frizzell, R. A., and Schultz, S. G.: Effects of uniiateral sodium replacement on sugar transport across in vitro rabbit ileum, Am.]. Physiol. 228: 854, 1975. Silverman, M.: Glucose transport in the kidney, Biochem. Biophys. Acta 457: 303, 1976. Nesbitt, R. E. L., Jr., Rice, P. A., Rourke, J. E., Torresi, V. F., and Souchay. A. M.: In vitro perfusion studies of
8.
9.
10. 11.
12.
13.
the human placenta. A newly-designed apparatus for extracorporeal perfusion achieving dual closed circulation, Gynecol. Invest. 1: 185, 1970. Rourke,]. E., Gould, L. V., Rice, P. A., and Nesbitt, R. E. L., Jr.: In vitro perfusion studies of the human placenta. Angioradiographic study of the maternal circulation, Gynecol. Invest. 4: 50, 1973. Nesbitt, R. E. L., Jr., Rice, P. A., and Rourke, J. E.: In vitro perfusion studies of the human placenta. III. The relationships between glucose utilization rates and concentration, Gvnecol. Invest. 4: 243, 1973. Feteris, W. A:: A serum glucose method without protein precipitation, Am.]. Med. Techno!. 31: 17, 1965. .Alexander, D.P., .A.ndrews, R. D., Huggett, .A. S. G., Nixon, D. A., and Widdas, W. F.: The placental transfer of sugars in the sheep: studies with radioactive sugar, ]. Physioi. i29: 352, 1955. Chenard, F. P., Danesino, V., Hartman, W. F., Huggett, A. S. G., Paul, W .. and Reynolds, S. R. M.: The transmission of hexoses across the placenta in the human and rhesus monkey,]. Physiol. 132: 289, 1956. Schultz, S. G.: Sodium coupled solute transport by small intestine. A status report, Am. J. Phys. 233: E249, 1977.