INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS AND MATERIALS RESULTS DISCUSSION REFERENCES

UNDER NORMAL CONDITIONS glucose in the maternal plasma is the source of glucose for uterine glucose consumption and provides all of the glucose used by the uteroplacenta (UP) and fetus (13, 15). Uterine arterial glucose concentration determines the net entry of glucose into the uterus and uteroplacenta from the maternal circulation, whereas independent changes in fetal arterial glucose concentration determine how much of the net uterine glucose uptake is consumed by the uteroplacenta or transferred to the fetus (12). Under normal conditions, however, fetal arterial plasma glucose concentration, and thus its effect on uteroplacental glucose consumption (UPGC), is directly related to the maternal plasma glucose concentration (G_M) (10, 11, 17). The uteroplacenta consists of uterine tissues (primarily endometrium and myometrium) and the placenta. In late-gestation pregnant sheep, blood flow distribution (24) and glucose clamp (12) studies indicate that the placenta probably accounts for most (70-80%) of uteroplacental glucose and oxygen metabolism. Fick principle measurements indicate that the principal products of uteroplacental glucose metabolism in pregnant sheep would include glycogen, lactate, fructose, and CO₂ (17, 27, 28, 36). However, the rates at which these metabolic products of uteroplacental metabolism are produced directly from glucose and the regulation of these metabolic rates by glucose supply have not been determined.

A variety of studies with tracer glucose also have been conducted in pregnant sheep. These experiments demonstrated lactate and fructose production from both maternal and fetal glucose in the maternal-placental-fetal unit (25, 28, 36), bidirectional transport of glucose across the uteroplacenta (18), and relative compartmentalization of fetal and uteroplacental lactate and fructose distinct from the maternal plasma (28, 36). None of these studies, however, quantified the production of lactate, fructose, or CO₂ from glucose metabolized in the uteroplacenta, in absolute terms or in relation to UPGC.

Previous data (17, 27) do, however, indicate that production rates for lactate, fructose, and CO₂ in the uteroplacenta are directly related to the rate of uteroplacental glucose consumption. Furthermore, if total uteroplacental lactate and fructose production were derived from UPGC, they would account for only about half of UPGC. Also, net UPGC, minus lactate and fructose production, could account for ~70% of net uteroplacental oxygen consumption, but only if the glucose consumed is completely oxidized. Also, rates of net uteroplacental oxygen consumption remain relatively constant despite marked differences in rates of UPGC (2, 12, 25). This observation indicates either an inverse relationship between UPGC and uteroplacental glucose oxidation, or a direct relationship between UPGC and uteroplacental glucose oxidation coupled with a reciprocal relationship between the oxidation of glucose immediately consumed by the uteroplacenta and the oxidation of other substrates.

Clearly there are several fundamental aspects of uteroplacental glucose metabolism that have not been determined. Therefore, we measured the rate of glucose consumption and the production rates of metabolic products of glucose by the uteroplacenta in late-gestation pregnant sheep under low (LG) and high (HG) G_M conditions. We infused U-[¹⁴C]glucose into the maternal circulation to provide glucose and tracer glucose to the uteroplacenta from the same precursor pool in the maternal plasma, and we compared tracer-derived metabolic rates with those measured by the Fick principle using established methods (8, 9, 13, 15, 28, 36, 37). We measured the net rate of UPGC, the net rate of uteroplacental oxygen consumption, and rates of glucose metabolism by the uteroplacenta into lactate, fructose, and CO₂ during both LG and HG conditions. These measurements were used to test the hypotheses that uteroplacental glucose metabolism is directly related to glucose supply and uptake, whereas uteroplacental oxygen consumption is relatively constant, demonstrating a reciprocal relationship between the oxidation of glucose just taken up by the uteroplacenta and the oxidation of other substrates.

	METHODS AND MATERIALS

TOP ABSTRACT INTRODUCTION METHODS AND MATERIALS RESULTS DISCUSSION REFERENCES

Animal care and use. Experiments were conducted in eight pregnant Columbia-Rambouillet mixed-breed sheep obtained from a commercial breeder (Nebeker, Santa Monica, CA), with known dates of gestation and numbers of fetuses. These sheep underwent surgery at 115-120 days gestation (term  148 days) under spinal anesthesia (6 mg tetracaine hydrochloride in 22.5 ml 10% wt/vol dextrose in water) and intravenous pentobarbital sodium sedation. This form of anesthesia produces no evidence of pain in the fetus or the mother, as determined every few minutes by the absence of reflex withdrawal from painful stimuli such as a needle prick. Using standard operating procedures, we placed 20-gauge polyvinyl catheters in accordance with the following scheme. Maternal catheters for infusion and sampling were placed into a maternal femoral vein and artery, respectively, via a groin incision, and another maternal sampling catheter was placed into the uterine vein of the pregnant uterine horn after exposure of the uterus via a midline laparotomy. Following a hysterotomy, fetal sampling catheters were placed into the common umbilical vein via direct insertion into one of the umbilical veins near the base of the cord, and into the lower fetal abdominal aorta near the takeoff of the umbilical arteries via a hindlimb pedal artery. A fetal infusion catheter was placed into each fetal femoral vein via a hindlimb saphenous vein. The fetal catheters exited the uterus through the sutured hysterotomy incision, and all catheters were collectively tunneled subcutaneously through a flank incision in the ewe and kept in a plastic pouch attached to the ewe's skin with stainless steel spikes. The catheters were filled and then flushed daily with sodium heparin solution (100 U in 0.9% wt/vol sodium chloride in water). Procaine penicillin (600,000 U) and gentamycin (80 mg) were given intramuscularly to the ewe prior to surgery, and ampicillin (500 mg) was left in the amniotic fluid at the end of the surgery. Following surgery, each animal received an intramuscular dose of ibuprofen. The ewes were allowed to recover for at least 5 days before starting experiments. They were fed an ad libitum diet of alfalfa pellets and water (measured daily), had free access to a mineral block, and received weekly intramuscular injections of a multivitamin mixture (B-complex vitamins; Vedco, St. Joseph, MO). The ewes were kept in standard carts that were cleaned as needed but at least daily in temperature-controlled rooms (18 ± 2°C), with 8 h of darkness each night. At least two sheep were always kept close to each other in the same room to decrease stress. At the end of the experiments, each ewe and fetus were killed with a rapid intravenous injection of pentobarbital sodium in 10% alcohol (Sleepaway; Fort Dodge Laboratories, Fort Dodge, IA). All animal care and study procedures were approved by the University of Colorado Health Sciences Center Institutional Animal Care and Use Committee. The University of Colorado Health Sciences Center Perinatal Research Facility is accredited by the Association for Assessment and Accreditation of Laboratory Animal Care, the United States Department of Agriculture, and the National Institutes of Health.

Study design. Two studies were performed on each animal, the first at ~126 days gestation and the second ~10 days later. The experimental protocols were the same for both studies. The two studies were done to test for gestational effects on uteroplacental and fetal glucose metabolism during a period of gestation when repeated studies in our lab for a variety of experimental protocols assume little or no developmental change in fetal and placental glucose metabolism. Food was removed from the ewe ~20 h prior to the first study. On the first morning of study, zero-time blood samples were taken for maternal and fetal plasma glucose, lactate, fructose, and amino acid concentrations, as well as blood oxygen content. The first or LG protocol was performed when the G_M was ~70% of normal (~3.9 mmol/l) and at least 0.06 mmol/l below the previous day's value. We did, however, allow for considerable variability in the actual G_M at the time of study. ³H₂O (Amersham, Arlington Heights, IL; ~35 mCi/ml 0.9% wt/vol sodium chloride in water) was infused at a primed (3.0 ml) constant rate (6 × 10⁶ dpm/min) through a fetal femoral vein catheter to measure umbilical and uterine blood flows by the steady-state transplacental diffusion technique (26). U-[¹⁴C]glucose (DuPont NEN, Boston, MA; 14 mCi/ml 0.9% wt/vol sodium chloride in water) was infused at a primed (3.0 ml) constant (4 × 10⁶ dpm/min) rate through a maternal femoral vein catheter to measure glucose metabolism. Two hours were allowed to reach steady state, after which blood samples were drawn simultaneously from the maternal femoral artery, uterine vein, umbilical vein, and fetal aorta at 120, 135, 150, and 165 min of infusion. These samples were processed for immediate measurement of blood oxygen saturation and oxygen carrying capacity; plasma glucose, lactate, fructose, and amino acid concentrations; plasma radioactive glucose, lactate, fructose, and CO₂ concentrations; plasma insulin concentration; and plasma ³H₂O concentrations. Fetal study period blood samples were replaced by equal volumes of heparinized maternal blood (drawn prior to infusion of tracer glucose).

At the end of this first LG protocol, food was returned to the mother. A maternal intravenous infusion of 50% wt/vol dextrose in water was started and adjusted in accordance with measurements taken every 30-60 min of maternal arterial plasma glucose concentration to achieve and maintain a maternal arterial plasma glucose concentration about twice normal. These procedures were based on modifications of our standard hyperglycemic glucose clamp protocol (10). The dextrose infusion rate at this time was continued overnight. The next morning, the maternal dextrose infusion rate was adjusted minimally over 1-2 h to stabilize and maintain approximately the same high-glucose steady-state glucose concentration achieved the night before. As in the LG studies, we allowed for variability in the actual glucose concentration at the time of study, but ensured that the study glucose value was at least 0.06 mmol/l greater than the prestudy value for each animal. The same experimental protocol was then followed as the day before with ³H₂O and U[¹⁴C]glucose tracers and sampling for labeled and unlabeled substrate concentrations. Zero-time samples were analyzed for radioactivity specific to glucose, lactate, fructose, and CO₂ to correct carryover radioactivity from the previous study. At the end of this second or HG protocol, the ewe was returned to normal conditions and diet. Both LG and HG protocols were repeated ~10 days later to test for developmental changes in the measured parameters. At the end of all studies, the ewe and fetus were killed, and an autopsy was performed to confirm catheter tip locations, weigh the fetus and placenta, and count the number of placental cotyledons.

Biochemical methods. Blood oxygen saturation and hemoglobin concentrations were measured in duplicate with a Radiometer OSM3 Hemoximeter (Radiometer, Copenhagen, Denmark). Plasma glucose and lactate concentrations were measured in duplicate with a Yellow Springs Model 2300 analyzer (Yellow Springs Instruments, Yellow Springs, OH). Plasma fructose concentration was measured in duplicate with a spectrophotometric enzymatic assay (Sigma, St. Louis, MO). Plasma radioactive glucose, lactate, and fructose concentrations were measured as previously described with ion exchange column chromatography (15, 28, 36). Glucose-, lactate-, and fructose-specific radioactivities were corrected to 100% recovery by means of control columns with known amounts of radioactive glucose or lactate added to control plasma samples. Blood ¹⁴CO₂ concentrations were determined as previously described (8, 9, 37) by decarboxylating 0.3 ml blood samples with 0.1 ml 1N hydrochloric acid under anaerobic conditions, trapping the liberated CO₂ overnight in a strong base.

Calculations. Blood oxygen content was calculated as the product of blood oxygen saturation × blood oxygen carrying capacity. Blood oxygen carrying capacity was determined as the product of blood hemoglobin concentration × oxygen carrying capacity of hemoglobin, 1.34 ml O₂/g hemoglobin. Umbilical and uterine blood flow rates were calculated according to the steady-state transplacental diffusion technique with ³H₂O tracer infused into the fetus at a constant rate (26). Plasma glucose-, lactate-, and fructose-specific radioactivities (dpm/mmol) were calculated as the ratio of plasma radioactive substrate concentration (dpm/ml) divided by plasma substrate concentration (mmol/ml). Transformation of plasma concentrations and specific radioactivities of substrates into whole blood values was done according to known blood-to-plasma ratios determined in our lab with control and experimental samples.

Net uterine uptake rates of glucose and radioactive glucose were calculated as the product of their uterine arteriovenous blood concentration differences and uterine blood flow rate (10). Net fetal uptake rates of glucose and radioactive glucose were calculated as the product of their umbilical venoarterial blood concentration differences and umbilical blood flow rate (18). Net uteroplacental consumption rates of glucose and tracer glucose were calculated as the differences between their uterine and fetal net uptake rates (13). Net uteroplacental lactate and radioactive lactate production rates were calculated as the sum of their net transfer rates into the fetus (umbilical venoarterial blood concentration difference × umbilical blood flow rate) and the mother (uterine venoarterial blood concentration difference × uterine blood flow rate) (36). Uteroplacental fructose and radioactive fructose production rates were assumed equal to net fetal uptake rates (umbilical venoarterial blood concentration difference × umbilical blood flow) (28). Total fetal and uterine ¹⁴CO₂ production rates were calculated as the product of umbilical arteriovenous or uterine venoarterial blood concentration differences and umbilical or uterine blood flow rate, respectively (8, 9, 37). Net uteroplacental ¹⁴CO₂ production rate was calculated as the difference between total uterine and total fetal ¹⁴CO₂ production rates. Net fetal ¹⁴CO₂ production rate was calculated by subtracting an estimated rate of fetal ¹⁴CO₂ derived from [¹⁴C]lactate and [¹⁴C]fructose that were produced in the uteroplacenta but metabolized in the fetus (18% of total fetal ¹⁴CO₂ production rate) from total fetal ¹⁴CO₂ production rate (25).

Maternal glucose disposal rate (GDR_M; µmol/min), equal to glucose rate of appearance during steady-state conditions, was calculated as total tracer glucose infusion rate (dpm/min) divided by the mean steady-state maternal arterial glucose specific radioactivity (SA_M; dpm/µmol) during the sampling period Nonuterine maternal glucose utilization rate (GUR_M; µmol/min) was calculated as [total tracer glucose infusion rate  net tracer uptake rate by the uterus (dpm/min)] divided by maternal arterial glucose specific radioactivity (dpm/µmol) where Uterine [¹⁴C]glucose arteriovenous difference was measured in dpm per milliliter, and uterine blood flow rate was measured in milliliters per minute.

Uterine glucose utilization rate (GUR_U; µmol/min) was calculated as net tracer glucose uptake rate (dpm/min) by the uterus divided by maternal arterial glucose specific radioactivity (dpm/µmol) Uteroplacental glucose utilization rate (GUR_UP; µmol/min) was calculated as net tracer glucose uptake rate by the uteroplacenta (dpm/min) divided by the uteroplacental glucose specific radioactivity (SA_UP; dpm/µmol), which was calculated as the average of maternal and fetal arterial glucose specific radioactivities Fetal glucose utilization rate (GUR_F; µmol/min) was calculated as net fetal tracer glucose uptake rate (dpm/min) divided by fetal arterial glucose specific radioactivity (SA_F; dpm/µmol) Uteroplacental lactate production (UPLPR; µmol/min) from glucose was calculated as the rate of [¹⁴C]lactate produced by the uteroplacenta (dpm/min) divided by the uteroplacental glucose specific activity (dpm/µmol). This expression was multiplied by 2 as 1 mol of glucose produces 2 mol of lactate Fetal glucose decarboxylation rate (i.e., the rate of CO₂ produced from glucose oxidation in fetal tissues; µmol CO₂/min) was calculated as net fetal ¹⁴CO₂ production rate (dpm/min) minus an estimated production rate of ¹⁴CO₂ in the fetus from [¹⁴C]lactate and [¹⁴C]fructose produced in the placenta but metabolized in the fetus (18% of total fetal ¹⁴CO₂ production rate) (25), divided by the fetal arterial glucose specific radioactivity (dpm/µmol CO₂). This expression was multiplied by six, because 1 mol of glucose produces 6 mol of CO₂ Uteroplacental glucose decarboxylation rate (i.e., the rate of CO₂ produced from glucose oxidation in uteroplacental tissues; µmol/min) was calculated as the uteroplacental ¹⁴CO₂ production rate (dpm/min) divided by the average of the fetal and maternal arterial glucose specific radioactivities (dpm/µmol). This expression was multiplied by six, because 1 mol of glucose produces 6 mol of CO₂ Fractional distributions of tracer [¹⁴C]glucose and radiolabeled products ([¹⁴C]lactate, [¹⁴C]fructose, ¹⁴CO₂) were calculated on the basis of absolute counts (e.g., 100 dpm of glucose could maximally produce 100 dpm of lactate). Fractional distributions of substrates (glucose, oxygen) and products (lactate, fructose, CO₂) were calculated on the basis of molar ratios (e.g., 1 mol of glucose could maximally produce 2 mol of lactate; 1 mol of glucose could maximally produce 6 mol of CO₂ and consume 6 mol of oxygen).

Data and statistical analyses. Duplicate measurements on each sample for each variable were averaged for a single mean sample value. Steady state in each sampling period was determined as less than ± 5% change of individual values over the sampling period with no consistent, significant trend to increase or decrease compared with the sampling period mean value. Sampling period mean values for each substrate or metabolic product were calculated as the average of the four samples. ANOVA with a means model of repeated measures that distinguishes between variability among subjects and variability among multiple measurements on the same subject over time was used to test for effects of gestational age, glucose concentration, and interaction between glucose concentration and gestational age (21, 23). Each analysis was based on four repeated measurements on each sheep (both LG and HG conditions at both first and second studies). If there was no significant gestational age effect for a given parameter, or if the gestational effect was very small (less than ± 10%), the results of the two studies were averaged for a single value for that parameter for that animal. Regression analysis was performed on the averaged data with a mixed-effects model that accounted for variability among sheep as well as the variability within sheep (21, 23). Values reported in the tables and text are the means among studies and animals of calculated as well as measured variables.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS AND MATERIALS RESULTS DISCUSSION REFERENCES

Eight pregnant sheep were studied, each on four occasions. The first two studies were conducted at 126 ± 1 (LG) and 127 ± 1 (HG) days of gestation, and the second two studies were conducted at 137 ± 1 (LG) and 138 ± 1 (HG) days of gestation. As shown in Table 1, absolute uterine and umbilical blood flow rates, which were within the normal range at each study, were not related to glycemic condition and did not change with gestational age between first and second study. Uterine and umbilical blood flow rates were constant during the sampling periods (see Fig. 4, bottom, showing the constant uterine and umbilical arteriovenous differences for ³H₂O). Fetal weights at the first study were based on fetal growth curves standard for this breed under usual dietary conditions in our lab (11). The expected weight gain during this period is ~1,100 g. Fetal weight-specific umbilical blood flow rates decreased with gestational age (24.4%, P < 0.05, for LG group; 32.6%, P < 0.05, for HG group; NS between LG and HG decrements) as a result of the estimated increases in fetal weights. At autopsy maternal and fetal weights were within normal ranges (Table 1), as were the number of cotyledons (66.8 ± 4.3) and the individual cotyledon weights (3.3 ± 0.3g).

Blood oxygenation. Blood hematocrit did not vary in mother or fetus according to glycemic condition (Table 2). There was a small increase between the 126- to 127-day and the 137- to 138-day studies of the maternal hematocrit in the LG group and the fetal hematocrit in the HG group. Also, maternal but not fetal blood O₂ content increased slightly between studies. These small gestational age differences were ignored in the final data analysis. Maternal and fetal arterial blood O₂ contents (Fig. 1) and the maternal (uterine) and fetal (umbilical) arteriovenous O₂ content differences (Fig. 2) did not vary significantly over the sampling periods. Fetal arterial blood O₂ content was 24% lower (P = 0.0001) in the HG group and was similarly decreased at both early and late studies. 3-5% of estimated fetal blood volume was replaced with maternal blood after each LG study, and another 3-5% was replaced after each HG study. There was no effect of glycemic condition or gestational age on net oxygen consumption by the uterus, fetus, or uteroplacenta.

View larger version (24K): [in this window] [in a new window]   Fig. 1.   Steady-state concentrations (means ± SE) of plasma glucose, blood oxygen content, plasma lactate, and plasma fructose during low (left) and high (right) glucose sampling periods. , Maternal artery; , uterine vein; , fetal artery; , umbilical vein.

View larger version (27K): [in this window] [in a new window]   Fig. 2.   Maternal uterine (UA-UV) and fetal umbilical (ua-uv) steady-state arteriovenous concentration differences (means ± SE) of plasma glucose, blood oxygen content, plasma lactate, and plasma fructose during low (left) and high (right) glucose sampling periods. None showed significant change over duration of sampling period. , Maternal artery; , fetal artery.

Concentrations and flux rates for glucose, lactate, and fructose. There were no statistically significant or metabolically large differences between the first and second studies for glucose, lactate, or fructose flux rates in the uteroplacenta (Tables 2-5). Therefore, results of the two studies were averaged for each animal for purposes of among-animal and between-glycemic-group comparisons. Steady-state concentrations (Fig. 1) and maternal (uterine) and fetal (umbilical) arteriovenous concentration differences (Fig. 2) were present for glucose, lactate, and fructose during the sampling periods. Arterial plasma glucose concentration in the HG group was 2.2-fold greater in the mother and 2.6-fold greater in the fetus than in the LG group. As a result of these glucose concentration differences, net uterine (+83%), fetal (+67%), and uteroplacental (+91%) glucose uptake rates all were greater in the HG group than in the LG group (P < 0.0001 for each). The percent increases in uteroplacental and uterine glucose uptake rates were not different, but both were greater than the increase in fetal glucose uptake (P < 0.05).

Arterial plasma lactate concentrations were greater in the HG group: +77% for the fetal values, which were significantly greater (P < 0.001) than the +19% for the maternal values. Uteroplacental lactate production rate was also greater in the HG group, 41% greater than the LG values (P < 0.01). Uteroplacental lactate production was partitioned more into the fetal than into the uterine circulation in the LG group (fetal/maternal net lactate uptake ratio = 1.32) and partitioned about equally in the HG group (fetal/maternal net lactate uptake ratio = 0.97). Compared with the LG condition, uteroplacental lactate production in the HG condition was greater for net uterine lactate uptake (+98%) than for net fetal lactate uptake (+44% for absolute uptake, P = 0.02; +41% for fetal weight-specific uptake, P = 0.04). This relatively greater increase in net lactate uptake rate by the uterine circulation compared with the fetal circulation from LG to HG condition was opposite of the relatively greater increase in fetal than maternal lactate concentrations from LG to HG condition. Uteroplacental lactate production rates represented about one-third of UPGC in both HG (29%) and LG (33%) conditions.

Fetal fructose concentration was 2.3-fold greater in the HG group than in the LG group. There was no net fructose uptake by the uterine circulation from the uteroplacenta. Net flux of fructose to the fetus from the uteroplacenta (uteroplacental fructose production rate) was not different between the two glycemic groups. Quantification of fructose flux rates was uncertain due to the very low extraction ratio of fructose by the umbilical circulation (e.g., the umbilical venoarterial fructose concentration difference divided by the fetal arterial plasma fructose concentration was only 0.6%). Fetal fructose uptake rates, and thus uteroplacental fructose production rates, represented 3% (LG) and 6% (HG) of uteroplacental glucose consumption.

Tracer concentrations, flux rates, and tracer-derived utilization and production rates. There were no statistically significant or metabolically large differences between the first and second studies for tracer-derived utilization and production rates; therefore, results of the two studies were averaged for each animal for purposes of between-animal and glycemic group comparisons (Tables 6-9). Steady-state conditions prevailed for [¹⁴C]glucose, ¹⁴CO₂, [¹⁴C]lactate, and [¹⁴C]fructose concentrations (Fig. 3) and for the arteriovenous concentration differences for these substances (Fig. 4). [¹⁴C]glucose infusion rates into the maternal circulation were not different between the two glycemic conditions. In contrast, the maternal (uterine) arterial glucose specific activity in the LG group was 3.1-fold greater than in the HG group. Thus, the maternal plasma glucose disposal rate in the HG group (26.6 ± 1.3 mmol · min¹ · kg¹) was 2.9-fold greater than in the LG group (9.2 ± 1.3 mmol · min¹ · kg¹).

Calculated fetal weight-specific glucose utilization rate was 80% greater in the HG group. This percent increase was greater than that for net glucose uptake (+67%), the difference approximately equal to the calculated contribution of fetal glucose production to fetal glucose utilization rate, although the glucose production value was not statistically significant.

Total tracer-derived uteroplacental glucose utilization rate was greater in the HG group than in the LG group (+2.2-fold); the values in both glycemic conditions were not different from those of the net uteroplacental glucose uptake rates. Similarly, tracer-derived net uteroplacental glucose utilization rate exclusive of uteroplacental lactate and fructose production rates was not different from Fick principle-derived net uteroplacental glucose uptake rate exclusive of uteroplacental lactate and fructose production rates.

View larger version (30K): [in this window] [in a new window]   Fig. 3.   Steady-state concentrations (means ± SE) of plasma [14C]glucose, 14CO2, [14C]lactate, and [14C]fructose during low (left) and high (right) glucose sampling periods. , Maternal artery; , uterine vein; , fetal artery; , umbilical vein.

View larger version (23K): [in this window] [in a new window]   Fig. 4.   Maternal uterine and fetal umbilical steady-state arteriovenous concentration differences for plasma [14C]glucose, 14CO2, [14C]lactate, [14C]fructose, and 3H2O during low (left) and high (right) glucose sampling periods. Mean arteriovenous differences were significant (P < 0.05) except for [14C]fructose, and none showed significant change over duration of sampling period. , Maternal artery; , fetal artery.

Calculated uteroplacental CO₂ production rates derived from direct uteroplacental glucose oxidation accounted for equal fractions of UPGC in the LG (0.16 ± 0.02) and HG (0.14 ± 0.03) groups, accounting for about one-third of uteroplacental glucose consumption that was not converted to lactate and fructose production. Uteroplacental glucose decarboxylation rate was directly and significantly related to UPGC (Fig. 5). Uteroplacental glucose oxidation was only ~16% of UPGC and did not vary between HG and LG groups (17 ± 2% LG, 15 ± 3% HG).

View larger version (14K): [in this window] [in a new window]   Fig. 5.   Mixed-effects regression analysis of uteroplacental glucose decarboxylation rate vs. uteroplacental glucose consumption rate for 8 animals studied in both low glucose (LG) and high glucose (HG) conditions (n = 16); y = 32.8 + 0.816(x) ± 60.5 SD, r = 0.80, P = 0.0001.

Direct fetal glucose oxidation (not including assumed rates of fetal oxidation of lactate and fructose that were derived from uteroplacental glucose metabolism) averaged 22.2 ± 2.4 mmol · min¹ · kg¹ in the LG group and increased significantly (P < 0.05) to 31.3 ± 3.6 mmol · min¹ · kg¹ in the HG group. It accounted for 74% of fetal glucose utilization in the LG group and 60% in the HG group.

Uteroplacental lactate production derived from UPGC was directly related to UPGC (Fig. 6) and was not different from total uteroplacental lactate production rate, representing from 29% (HG) to 33% (LG) of UPGC (values not different). Net uteroplacental lactate production was 67% greater in the HG than in the LG group. Lactate production into the fetal circulation was 31.2% greater than into the uterine circulation in the LG group; in the HG group, however, the fetal-to-maternal distribution ratio was 0.97. Lactate production by the uteroplacenta accounted for similar fractions of net UPGC between glycemic groups, 0.33 and 0.29 for LG and HG groups, respectively. Fetal net lactate uptake rates accounted for similar fractions of UPGC between the two glycemic conditions (0.13 LG, 0.10 HG). Lactate specific activity was not different between the umbilical vein and artery or between the uterine vein and artery.

View larger version (14K): [in this window] [in a new window]   Fig. 6.   Mixed-effects regression analysis of uteroplacental lactate production rate vs. uteroplacental glucose consumption rate for 8 animals studied in both LG and HG conditions (n = 16); y = 37.7 + 0.423(x) ± 26.5 SD, r = 0.90, P = 0.0001.

Fetal fructose utilization rates were not different from net fructose uptake rates in the LG group, but were considerably greater than uptake rates in the HG group. There was considerable variability, however, in these values, resulting in no significant difference between glycemic groups for either tracer-derived or Fick principle-derived uteroplacental fructose production rates.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS AND MATERIALS RESULTS DISCUSSION REFERENCES

We found a surprisingly small contribution of direct glucose decarboxylation to carbon dioxide production (uteroplacental decarboxylation rate was 15-18% of UPGC), with uteroplacental glucose oxidation accounting for only 23 (LG) to 34% (HG) of uteroplacental oxygen consumption. Thus most of the oxygen consumed by the uteroplacenta appears to be used to oxidize carbon other than from the relatively immediate metabolism of glucose just taken up by the uteroplacenta. If this is true, then it is reasonable to assume that there is considerable turnover of carbon in the placenta, which would rapidly dilute tracer glucose metabolism to a large extent by carbon released from other unlabeled sources, such as glycogen, lipids, and proteins. A very small amount of lactate derived from uteroplacental fructose production is oxidized in the uteroplacenta (28), and previous Fick principle measurements have defined only very small uteroplacenta uptake rates for free fatty acids and ketoacids (20, 30). Recent data have defined net uteroplacental consumption rates for amino acids in the ovine placenta near term (4). Assuming, very roughly, an average 25% fractional oxidation rate (3, 6, 19, 29, 35) for those amino acids consumed in net by the uteroplacenta (serine, glutamate, valine, leucine, and isoleucine), another 1-3% of uteroplacental oxygen consumption could be accounted for by amino acid oxidation in the uteroplacental tissues. Also, we based our calculations of uteroplacental glucose oxidation on the assumption that with uniformly labeled glucose, all six carbons would be oxidized to ¹⁴CO₂ during the 165 min of our study. This assumption is reasonably based on other studies in the fetus (8) and adult (39) and also on the lack of evidence for gluconeogenesis in the uteroplacenta (1). Such assumptions, however, have not been tested, and there may indeed be an incomplete oxidation of all glucose carbons during the period of our study. It also is important that the flux measurements and metabolic relationships that we measured are unique only to the conditions at the time we performed our studies. Some fluxes might be different (especially fetal gluconeogenesis) if the prevailing glycemic conditions were maintained for longer periods.

Consistent with previous in vivo studies in the sheep (27), uteroplacental oxygen consumption was relatively large, averaging 63% of fetal oxygen consumption on an absolute basis, and nearly 10-fold more than the fetus on a weight-specific basis. Uteroplacental oxygen consumption was not different between LG and HG conditions, yet glucose oxidation accounted for a 50% greater fraction of uteroplacental oxygen consumption in the HG period compared with the LG period. These results demonstrate a reciprocal relationship between the oxidation of glucose just taken up by the uteroplacenta and other substrates to account for the maintenance of uteroplacental oxygen consumption, a phenomenon previously observed in fetal sheep (8, 9).

Fetal arterial and umbilical venous blood oxygen contents were significantly lower in the HG group. These results are consistent with many previous observations (8, 31, 32, 34), and have been ascribed to an increased rate of oxygen consumption by the fetus due to hyperglycemic stimulation of metabolic rate (8, 31, 32). Although fetal oxygen extraction was significantly increased in the hyperglycemic animals in the present study, there were variable counterbalancing increases in umbilical blood flow. Thus, we did not observe a significant increase in fetal oxygen consumption rate. This result also may represent an insufficient number of animals for the accuracy of the blood oxygen content measurements with the Radiometer OSM III Hemoximeter and the variable glucose concentrations achieved, which produced variable effects of increased fetal glucose concentration on oxygen consumption. Also, umbilical venous oxygen content was lower in the HG group. This would represent either an increased consumption of oxygen by the uteroplacenta, which we did not observe, nor have others (32), or evidence of the established low oxygen diffusing capacity of the ovine uteroplacenta (38). A further contribution to the lower fetal blood oxygen contents in the HG group may have been the greater volume of fetal blood replaced with lower oxygen affinity maternal blood during the sampling periods. This volume was 6-10% of estimated fetal blood volume at the end of the HG studies compared with 3-5% at the end of the LG studies. At most, such potential changes in hemoglobin-oxygen affinity would have accounted for only a small fraction of the ~24% decrement in fetal arterial oxygen content at the end of the HG studies. Because fetal and uteroplacental oxygen consumption rates were not different between the two groups, we doubt that the lower oxygen contents in the HG animals had independent effects on uteroplacental glucose metabolism, although further investigation of this phenomenon is warranted.

Uteroplacental lactate production rate was greater in the HG group and was partitioned more into the uterine circulation than into the umbilical circulation of the fetus, in contrast to the differences in maternal and fetal lactate concentrations, which were greater in the fetus in the HG group. The direction of lactate transport across membranes via the monocarboxylate transporter is determined largely by the transmembrane lactate concentration gradient (22, 33). One explanation for the greater relative distribution of uteroplacental lactate production into the maternal circulation in the HG group, therefore, is that the intratrophoblast-to-maternal plasma lactate concentration gradient was greater than the intratrophoblast-to-fetal plasma lactate concentration gradient. This and other possible mechanisms need experimental verification.

Uteroplacental lactate production derived from UPGC was not different from total uteroplacental lactate production, indicating that all of uteroplacental lactate production was derived from UPGC. This conclusion is supported by the lactate and glucose specific activity relationships. Lactate specific activity did not change across the uterine or umbilical circulation. Furthermore, glucose specific activity was the same in both maternal and fetal plasma. This evidence indicates that there was no significant lactate produced by uterine tissues from nonlabeled substrate sources. Thus, even though the uterus takes up other substrates, including amino acids such as alanine (6) and glutamine (4), and keto acids such as -hydroxybutyrate and acetoacetate, their contribution to net uteroplacental lactate production over a broad range of uteroplacental glucose supply and consumption was negligible (20, 30). In contrast, the mean specific activity of lactate in both maternal and fetal plasma was much lower than that of glucose, by 69 (LG) and 60% (HG) in fetal lactate and by 72 (LG) and 54% (HG) in maternal lactate. This indicates that a considerable fraction of circulating maternal and fetal lactate was derived from nonglucose precursors in both fetus and mother. The relatively large source of lactate from nonglucose precursors in the fetus found in this study is consistent with that estimated by Sparks, et al. (36), who used a fetal infusion of tracer lactate. In that study, ~50% of fetal lactate turnover was estimated to result from fetal production of lactate, and of that value, half came from glucose and half from nonglucose precursors. The lower fractional production of lactate from nonglucose precursors in the HG condition also supports the additional observations by Sparks, et al. (36) that demonstrated a reciprocal relationship between production of lactate from glucose and from nonglucose precursors.

The fetal plasma concentration of fructose increased markedly during the high glucose period, consistent with previous observations (25). This could be due to increased fructose production, which has been demonstrated from low to normal increases in uteroplacental glucose consumption (28), from a limited capacity of the fetus to metabolize fructose (28), or from competitive inhibition of fructose consumption in the fetus by the high glucose concentrations. We did not measure a significantly increased rate of fructose production by the uteroplacenta or fructose consumption by the fetus, although the results trended in that direction. Nevertheless, we believe that these fluxes were increased, just not measurable because of the very small plasma concentration differences at very high concentrations across the umbilical circulation. The analytical method for determining plasma fructose concentration cannot distinguish <1% differences. Even a 1% increase in the umbilical venoarterial fructose concentration difference could lead to an increased but unmeasurable rate of fructose production (~5.0 mmol · min¹ · kg¹), but is more than twice "normal" fetal fructose uptake and utilization rates and possibly greater than the fetus could metabolize.

Perspectives