Placental Transfer of Lactate, Glucose and 2-deoxyglucose in Control and Diabetic Wistar Rats

Placental transfer of lactate, glucose and 2-deoxyglucose was examined employing the in situ perfused placenta. Control and streptozotocin induced diabetic Wistar rats were infused with [U14C]-glucose and [3H]-2-deoxyglucose (2DG). The fetal side of the placenta was perfuseci with a cell free medium and glucose uptake was calculated in the adjacent fetuses. Despite the 5-fold higher maternal plasma glucose concentration in the diabetic dams the calculated fetal glucose metabolic index was not significantly different between the 2 groups. Placental blood flow was reduced in the diabetic animals compared with controls but reduction of transfer of [U14C]-glucose and [3H]-2-deoxyglucose and endogenously derived [14C]-Lactate to the fetal compartment, could not be accounted for by reduced placental blood flow alone. There was no significant net production or uptake of lactate into the perfusion medium that had perfused the fetal side of the placenta in either group. The plasma lactate levels in the fetuses adjacent to the perfused placenta were found to be higher than in the maternal plasma and significantly higher in the fetuses of the diabetic group compared with control group. In this model the in situ perfused placenta does not secrete significant quantities of lactate into the fetal compartment in either the control or diabetic group.


INTRODUCTION
Lactate is a three carbon molecule derived from anaerobic metabolism that occurs in most cells.
Classically lactate has been considered a product of anaerobic glycolysis or a substrate for gluconeogenesis or glycogenesis in liver. Lactate also has an important role in carbohydrate metabolism by virtue of the rapid equilibrium that exists between lactate and pyruvate, which allows lactate to enter the citric acid cycle. It therefore represents a major energy source.
Under aerobic conditions lactate levels within tissues are usually low due to its rapid turnover rates. [1][2][3][4] The exception being in normal mammalian pregnancies where lactate concentrations are higher in fetal than in the maternal circulation, in humans, [5,6] sheep, [7,8] guinea pigs [9] and rats. [1] The elevated levels of this metabolite in the fetal circulation were shown to originate from the placenta in guinea-pigs using an in vitro placental preparation. [9] This has also been demonstrated in sheep in in-vivo experiments at term, [7,81 *Corresponding author. 114 C.R. THOMAS et al.
but not at midgestation. [111 This led to the assumption that the elevated fetal plasma lactate levels could be derived from placental anaerobic glycolysis. In diabetic pregnancies fetal plasma lactate concentrations have been reported to be significantly elevated compared with those of non-diabetic pregnancies. [12] Additionally placental glycogen stores are significantly increased [13] compared with those of non-diabetic pregnancies.
Thus the higher fetal lactate concentration in the presence of maternal diabetes may be of placental origin, especially as fetal pO 2 values of the umbilical vessels from diabetic patients were reported to be within the normal range. [12] The aim of these studies was to determine whether circulating maternal glucose could be the source of lactate in the fetal circulation of normal and diabetic pregnant rats. This was achieved by examining the transfer of glucose and lactate derived from glucose across the insitu perfused placenta. Fetal plasma concentrations of glucose and lactate as well as tissue glucose uptake were also estimated.  Keynes) and rat insulin standard used in the radioimmunoassay from Nova Biolabs Ltd. (Basingstoke) and the rabbit anti-guinea pig antibodies were from Immunodiagnostics Ltd. (Tyne and Wear).

MATERIALS AND METHODS
Streptozotocin, bovine serum albumin (Fraction V), D-glucose, antipyrine, perchloric acid, formic acid and all other chemicals used in the enzymatic and radiometric assays were obtained from BDH Chemicals Ltd. (Poole, Dorset) and were of analytical grade.

Animal Experimentation
Wistar rats (200-250g; B and K Universal Ltd. Hull) were mated and day 0 of pregnancy was designated as the day a vaginal plaque was noted. On day 4, ten rats were made diabetic by a single intra-peritoneal (i.p.) injection of streptozotocin (40mg/kg). Rats were maintained on standard breeding rat chow throughout gestation. This diet consisted of 31.5% protein, 43.5% carbohydrate, 4.0% fats, 12.2% dietary fibre and 2.5% supplementations (vitamins, minerals and amino acids). On day 20 of gestation, rats were anaesthetized with Intraval (i.p.) at a dose of 70mg/kg for controls and 60mg/kg body weight for diabetic rats. Once anaesthetized, the rat was placed on a thermo-regulated heating pad where the rat body temperature was maintained at 38C by feedback from a rectal probe.

Cannulation of Maternal Vessels
Experimentation procedures carried out on the rats were previously described. [14] In brief, the left maternal external jugular vein was cannulated to enable the infusion of radioactive cocktail and antipyrine in saline. The right carotid artery was cannulated for maternal blood sampling and the monitoring of blood pressure via a side line connection. Once the neck vessels were cannulated and infusion commenced a laparotomy was performed, a fetus was exteriorised to expose the umbilical vein and artery that were then cannulated prior to the removal of the fetus. The placenta was perfused via the artery at a rate Of 0.5ml/min for 30min, during which 10 threeminute (1.5ml) perfusate fractions were collected from the umbilical vein into chilled pre-weighed tubes. The perfusion fluid consisted of Krebs-Ringer bicarbonate buffer (pH 7.4) supplemented with 30mg/L Dextran 40 and 5g/L bovine serum albumin. D-glucose (5mmol/1 for control, 20mmol/1 for diabetic animals) and D-lactate (10mmol/1 for both control and diabetic animals) were added to the perfusion fluid to mimic fetal glucose and lactate levels (see Tab. II). A measure of a satisfactory experiment was that 96-100% of the inflowing perfusion fluid was recovered after a single passage through the placenta and that the adjacent fettlses were alive until removed from their respective placentae.

Maternal and Fetal Blood Sampling
During the 30 minute perfusate collection, 6 maternal blood samples (0.6ml) were taken at 6min intervals and divided equally between chilled pre-weighed heparinized test tubes and those containing 10% perchloric acid. At the end of the experiment, the remaining fetuses in utero were sequentially removed. Fetal blood samples were obtained from cut axillary vessels with a heparinized Pasteur pipette and the samples from each litter pooled. A final maternal blood sample was then taken. All plasma samples were aliquoted prior to storage at -20 C.  [16] using microcolumns (0.5cm 3 cm) of ion exchange resins AG 50W 8X (cationic) and AG 1 8 (anionic) in tandem. Recoveries for [14C]-glucose, [3H]-2-deoxyglucose and [14C]-lactate from these microcolumns were between 95-98%. Optiphase Hisafe 3 scintillation fluid was added to the collected eluates and counted for /3-emission in a dual channel programme on the LKB 1219 Rackbeta Spectral liquid scintillation counter. Quencing and spillover of the high-energy isotope counts into the low-energy counting channel were determined and corrected for by the channel ratio method, using an external standard and two "curves" of CC14-quenched [3H] and [14C]n-hexadecane standards.

Analysis of Samples
Maternal and fetal plasma and perfusate were analysed for glucose using the glucose oxidase method (YSI 23 AM analyser, Yellow Springs, OH, USA). Antipyrine was measured in all maternal and perfusate samples by the method of Brodie et al. [17] Insulin measurements on maternal and fetal plasma was by an in house radioimmunoassay [18] using an insulin antibody raised in guinea pigs and a rat insulin standard. 2-deoxyglucose-6-phosphate (2DG6P) levels were determined in fetal and placental tissues by the method of Ferre et al. [19] Calculations Maternal (M) to perfusate (P) transfer was calculated as the individual ratios of counts in the perfustae (dpm/ml) to that in the maternal plasma (dpm/ml). The values were first averaged from each rat. A grand mean for each group was then calculated. Glucose uptake for fetal and placental tissues were estimated using the glucose metabolic index (GMI) calculated based on the ratio of

RESULTS
Maternal weight, litter size and fetal weight were significantly reduced in the diabetic rats compared with normal rats. The converse was seen with the placentae in the two groups (Tab. I).
The administered dose of streptozotocin increased maternal glycaemia 5 fold (Tab. II) and decreased the maternal plasma level of insulin by 64% (Tab. II). Fetal glucose concentrations in the diabetic group were similarly elevated, however the decrease in fetal plasma insulin level was only 38%. Maternal plasma lactate concentrations were similar in the two groups. Fetal plasma lactate concentrations were significantly higher in the diabetic animals compared with the controls but more striking was the difference in concentration between the maternal and fetal levels in both diabetic and control animals (Tab. II). The placentae were perfused with 10mmol/1 lactate and the perfusate lactate concentation was estimated in each perfusate aliquot. There was a non significant net production of 0.88 0.37 mol/minute for the control animals and a non significant net uptake of 0.33 +0.34mol/minute in the diabetic animals. These values were not significantly different from each other. Fetal blood was collected at the end of the experiment from the remaining fetuses when the blood flow to the placenta may have been compromised, possibly resulting in fetal hypoxia and elevating the lactate concentration further. Antipyrine, a diffusable marker, was used as an indirect measure of maternal placental blood flow. The placental transfer of antipyrine, glucose and lactate was expressed as a ratio of that in the perfusate to that in the simultaneous maternal plasma sample (P:M), (Tab. III). The antipyrine ratio was reduced to 59% in the diabetic group compared to the control group, indicating a decrease in uterine blood flow in the former.       The calculated glucose metabolic index of control and diabetic placentas and fetal carcasses lactate.
The quantity of [3H]-2-DG-6-P present in the placentae, whole fetuses and fetal livers of diabetic group was significantly less compared with the controls group (p K0.05, p<0.0005, p 0.0005, respectively; Fig. 1). However, when the GMI was calculated fetal uptake of glucose was not significantly different between control and diabetic animals. The calculated GMI for the fetal livers and placentas were significantly increased in diabetic group compared with the controls (Fig. 2).

DISCUSSION
The main finding in this rat model was that the elevated fetal lactate concentrations observed in control and diabetic animals were not derived from the placenta, as there was no significant uptake or secretion of lactate from or into the perfusion medium. This was a somewhat unexpected finding. In vivo sheep experiments have shown that in mid gestation there is no secretion of lactate by the placenta to the fetal circulation, but at term lactate secretion to the fetus is substantial. [7,11] In our model we only examined placental lactate production and transfer to the fetal compartment. Aldoretta and Hay [2] have examined lactate production derived from glucose in late gestation in sheep under hypoglycaemic and hyperglycaemic states. Less lactate derived from glucose was secreted into the fetus in the hyperglycaemic than in the hypoglycaemic sheep. This may be due to the higher fetal plasma lactate concentration observed in the hyperglycaemic sheep, since direction of lactate transport is determined by the transmembrane lactate concentration. [21] The lactate concentration in the medium perfusing the fetal side of the placenta in our experiments was 10mmol/1 considerably higher than the maternal concentration. This gradient effect may also account for lactate secretion when placental slices are incubated without added lactate in the incubating medium. [22] Approximately 40% of ovine fetal lactate is not derived from glucose and alanine is the likeliest source. The production of [14C] lactate from L-[14C] alanine has been examined by Palcin et al. [22]  fetuses adjacent to the perfused placenta. Significant quantities of fetal lactate may thus be derived from fetal rather than from placental alanine deamination. Glycogen levels have been shown to be increased in the diabetic rodent placenta. [13] Lactate production from placental glycogen synthesised from infused universally labelled glucose during the 30-minute experiment would have caused a gradual rise in the P:M [14C]-lactate ratio. However, this was not ob-served. The observed placental secretion patterns in different models and at different gestational ages would suggest that there is a control mechanism determining lactate secretion to the fetal and or maternal circulation. Lactate is a universal alternative substrate to glucose for tissues including the brain. [23,24] Thus maintaining the plasma fetal lactate level above the maternal concentration may provide an alternative fetal nutrient.
The observed elevated fetal lactate levels of the diabetic dams compared with controls were not due to maternal to fetal placental transfer since the P:M [14C]-lactate ratio, together with the P:M ratio of [U4C]-glucose and [3H]-2-DG, were significantly lower compared with control animals. Additionally the total maternal plasma lactate levels were similar in the two groups. In this model the elevated fetal lactate values in either group have not been shown to be of placental origin.
The documented presence of very low enzyme activities of the citric-acid-cycle [25] in fetal hepatocytes may account for the presence of elevated fetal lactate concentrations compared with the mother. The significantly higher fetal lactate levels in the presence of maternal diabetes were probably secondary to reduced fetal insulin levels. The combination of the reduced P:M ratio of [U4C]-glucose, and a higher F :M ratio in diabetic compared with control animals, suggests that against the background of maternal hyperglycaemia there was reduced fetal glucose uptake. This is supported by the lower deposition of 2-[3H]-DG-6-P observed in the fetuses and placentas of diabetic dams. Despite this decrease in 2-[3H]-DG-6-P deposition, total glucose uptake from calculated GMI was increased in the placenta. The higher F:M ratio of the specific activities of glucose in the maternal and fetal circulation in the diabetic group (0.86_+0.05) compared with controls (0.41 _+ 0.03) suggests that the fetuses of the diabetic dams were only able to achieve glucose uptake similar to the control fetuses at hyperglycaemic glucose concentrations. The GMI of the whole fetus in the two groups were similar although fetal liver GMI, an insulin independent tissue, was increased. This could be a protective mechanism whereby the redu'ced fetal plasma insulin levels limits glucose metabolism in insulin sensitive tissues and prevents excessive somatic growth. This is supported by the observation of reduced fetal myocardial and skeletal muscle insulin sensitive Glut-4 transporter expression in chronically hyperglycaemic sheep. [26] Fetal insulin synthesis would not have been affected by streptozotocin as the drug was administered to the mother prior to the implantation of the conceptus (day 5). Reduced fetal insulin secretion in the diabetic group may, therefore, be a direct result of glucotoxicity, a phenomenon also observed in sheep with chronically elevated maternal plasma glucose levels. [27] The observed increased diabetic placental glucose uptake is in agreement with Thomas and Lowy [28] but differed in magnitude. This may be related to two factors, the rats in this study were only moderately diabetic compared with the previous study, and they were maintained on a diet optimal for breeding which had been enriched in protein (20.5% w/w) compared to the standard diet used in the previous study. CONCLUSION Placental glucose transport is less efficient in diabetic animals, but is compensated for by the fivefold higher maternal glucose concentration, resulting in an increased delivery of glucose to the fetal compartment.
[3H]-2DG-6-P deposition was significantly reduced in fetuses of diabetic dams. As the result of the five-fold higher fetal plasma glucose concentration in the diabetic model, the total fetal carcass glucose uptake, as calculated by GMI, was similar in the 2 groups. This relatively reduced fetal glucose uptake may be due to the low insulin levels in the diabetic animals. The P:M ratios observed for tracer lactate, glucose and 2-deoxyglucose indicate that the placenta will transport lactate. However, the placenta did not secrete significant quantities of lactate in this experimental model. The elevated fetal lactate concentrations are therefore a" primary product of fetal metabolism. This metabolic pathway is enhanced in the diabetic animal resulting in higher plasma fetal lactate levels.