Journal of Applied Physiology
Vol. 81, No. 6, pp. 2393-2398, December 1996
CONTROL OF BREATHING, CIRCULATION, AND TEMPERATURE

Suppressive action of endogenous adenosine on ovine fetal nonshivering thermogenesis

Karen T. Ball, Tania R. Gunn, Peter D. Gluckman, and Gordon G. Power

Center for Perinatal Biology, School of Medicine, Loma Linda University, Loma Linda, California 92350; and Department of Pediatrics, School of Medicine, University of Auckland, Auckland, New Zealand

ABSTRACT

INTRODUCTION

METHODS

RESULTS

DISCUSSION

ACKNOWLEDGEMENTS

FOOTNOTES

REFERENCES

ABSTRACT

Ball, Karen T., Tania R. Gunn, Peter D. Gluckman, and Gordon G. Power. Suppressive action of endogenous adenosine on ovine fetal nonshivering thermogenesis. J. Appl. Physiol. 81(6): 2393-2398, 1996.Nonshivering thermogenesis is not initiated when the fetal sheep is cooled in utero but appears to require the removal of an inhibitor of placental origin at birth. To test whether adenosine is such an inhibitor, we examined the effect of the adenosine antagonist theophylline on the initiation of nonshivering thermogenesis during sequential cooling, ventilation, and umbilical cord occlusion in utero. Theophylline (18 mg/kg bolus and 0.6 mg ^· kg^{1 ·} min¹ thereafter) was infused for 90 min before and 90 min after cord occlusion. Theophylline enhanced the nonshivering thermogenic free fatty acid (FFA) and glycerol responses before cord occlusion, raising FFA concentrations 99% to 415 ± 60 µeq/l (P < 0.01) and glycerol levels 87% to 526 ± 135 µmol/l (P < 0.05). These FFA (P < 0.001) and glycerol (P < 0.05) concentrations were significantly greater than the corresponding period during the birth-simulation control. Umbilical cord occlusion did not alter FFA levels but induced a 41% rise in glycerol concentrations to 774 ± 203 µmol/l (P < 0.05). The increases in nonshivering thermogenic indexes after the administration of the adenosine-receptor antagonist suggest that the quiescent state of ovine fetal brown adipose tissue may result, in part, from the tonic inhibitory actions of adenosine and that a decrease in adenosine concentrations enhances nonshivering thermogenesis. However, the further rise after umbilical cord occlusion suggests that at least one other inhibitor of placental origin inhibits nonshivering thermogenesis before birth.

brown adipose tissue; placenta; A₁-receptor antagonist

INTRODUCTION

WHEN THE FETAL SHEEP is cooled in utero, it produces little heat from nonshivering thermogenesis in brown adipose tissue, as evidenced by the failure of plasma free fatty acids (FFA) and glycerol concentrations to increase (16). The failure of this thermogenic response is observed despite the capability of ovine fetal brown adipocytes to respond to thermogenic stimuli if removed from the body (12) and despite the ability of prematurely delivered and term-delivered lambs to produce significant amounts of additional heat from nonshivering thermogenesis in brown fat (1). These findings indicate that brown adipose tissue is competent but unresponsive before birth.

The stimulus, or combination of stimuli, that may account for the initiation of nonshivering thermogenesis within minutes of birth is unknown. Although a number of potential stimuli have been tested and found to be necessary, none has proved sufficient alone. For example, cooling and increased oxygenation of the fetal sheep (16) and an intact sympathetic nervous system (2) are necessary for thermogenesis but fail to evoke responses comparable to those seen after birth. Only after the umbilical cord has been occluded do the thermogenic responses increase appreciably (17). Such observations suggest that the placenta produces a thermogenic inhibitor(s) that circulates before birth and tonically suppresses nonshivering thermogenesis. Once the umbilical cord is occluded, the inhibitor can no longer enter the fetal circulation, and nonshivering thermogenesis commences.

Our previous observations with prostaglandin E₂ and indomethacin during the simulation of birth in the fetal sheep clearly show an inhibitory role for placental prostaglandins in nonshivering thermogenesis (9). Indomethacin administered before cord occlusion enabled prompt responses to increased oxygenation and cooling alone; the rise in nonshivering thermogenesis that would generally occur on umbilical cord occlusion was minimized. Furthermore, this response was inhibited by infusion of prostaglandin E₂ (9). Because cord occlusion induced a further small rise in thermogenic indexes after indomethacin treatment, however, there was the suggestion of multiple intrauterine inhibitors of nonshivering thermogenesis in the fetal sheep.

Adenosine is a compound that is known to be released by the placenta (24), to be present in high concentrations in fetal plasma (19), to have a short half-life in circulating blood (5), and to inhibit catecholamine-stimulated lipolysis in brown adipose tissue in vitro (29). That adenosine is a potential candidate for one of the placental inhibitors of nonshivering thermogenesis (16, 17) is suggested by the effects of the infusion of N⁶-(L-2-phenylisopropyl)-adenosine (PIA), a stable, long-acting, and extremely potent agonist, selective for adenosine A₁-receptors on fat cells during the simulation of birth in utero. Low and medium doses of PIA administered after umbilical cord occlusion and during maximal nonshivering thermogenesis induced a dose-dependent fall in fetal core temperature, FFA and glycerol concentrations, and oxygen consumption (4). Thus we hypothesized that the high intrauterine levels of adenosine would inhibit nonshivering thermogenesis in utero, but the decreased adenosine levels on removal of the umbilical circulation would therefore enable the initiation of thermogenesis. In these experiments, we examined this hypothesis by simulating the birth of fetal sheep in utero and administering the adenosine antagonist theophylline before and during umbilical cord occlusion. Results are compared with a control study undergoing simulated birth without adenosine antagonist administration.

METHODS

Surgical and Postoperative Procedures

After the surgery, the ewes were housed in metabolic cages at a constant temperature (18°C) and humidity (50%) and given free access to concentrates and water. The experiments were performed while the ewe stood quietly in the cage. All studies were performed 24-72 h after surgery on fetuses with normal arterial blood pH (7.34), arterial PO₂ (Pa_O2; 18 Torr), and arterial PCO₂ (Pa_CO2; 55 Torr).

Experimental Procedures

Fetal temperature measurements were made by connecting the thermistors to a unit capable of linear temperature measurement from 32-43°C, with a resolution of 0.01°C. Recordings of fetal arterial blood pressure, tracheal pressure (both corrected for amniotic pressure), amniotic pressure, and heart rate were monitored by using calibrated transducers and a Gould 200 recorder (Gould Instruments). Recordings were begun 1 h before experimentation and continued until the termination of the study.

Simulation of birth in utero. The simulation of birth in utero was undertaken as described previously (16). Briefly, cooling was induced by passing iced water through the fetal thoracic cooling coil at a rate of 20-40 ml/min. This rate was adjusted to cause a fall in fetal core temperature of ~2°C in the first 60 min. In protocol 1, to maintain a constant sensory stimulus, the rate of cooling was adjusted to maintain the fetal core temperature at the level reached after 60 min of cooling. However, in protocol 2, the rate of cooling during the period of cold exposure was sustained throughout the rest of the study. This was done so that changes in body temperature would provide an additional index of thermogenic activity. After the tracheal cannula was drained of fluid, ventilation with oxygen saturated with water vapor was begun. The ventilatory circuit consisted of a respiratory pump (model 607, Harvard), a 1-liter spirometer to measure oxygen uptake by the lungs, a system to maintain an end-expiratory pressure of a few mmHg, appropriate one-way valving, and a CO₂ absorber (15). Effective ventilation was typically maintained with a respiratory frequency of 20-30 breaths/min, and peak-inspiratory and end-expiratory pressures were kept 20-40 and 3-10 mmHg higher than amniotic fluid pressure, respectively. Fetal Pa_O2 rose to 50-270 Torr, whereas CO₂ levels and pH were unchanged. Umbilical cord occlusion was induced in protocol 1 by pulling the loop tight around the umbilical cord to stop all umbilical blood flow. A simultaneous increase in mean arterial blood pressure of ~10 mmHg and a 25 beats/min rise in heart rate indicated effective occlusion of the cord.

Both an experimental and a control protocol were performed.

Protocol 1: administration of theophylline during birth simulation (n = 6). The day following surgery, after a 30-min control period (from 30 to 0 min) the fetus was cooled (0-300 min). The rate of cooling was adjusted to cause a fall in fetal core temperature of ~2°C in the first 60 min of cooling and was adjusted thereafter to maintain the fetal core temperature at the level reached after 1 h of cooling. Sixty minutes after cooling commenced, the fetus was ventilated with oxygen (60-300 min). The nonspecific adenosine antagonist theophylline (1,3-dimethylxanthine, 18 mg/kg bolus, then 0.6 mg ^· kg^{1 ·} min¹) was administered intravenously to the fetus 60 min later and continued until the end of the study (120-300 min). The umbilical cord was occluded 90 min after commencing the theophylline infusion, and the responses were followed for a further 90-min period (210-300 min), with the fetus isolated from the placenta.

Protocol 2: birth simulation control (n = 8). One to three days after surgery, after a 30-min control period (from 30 to 0 min), the fetus was cooled (0-300 min). The rate of cooling was adjusted to cause a fall in fetal core temperature of ~2°C in the first 60 min of cooling, and this rate of cooling was sustained throughout the rest of the study. The different methods of cooling in the experimental and control protocols resulted in comparable responses of fetal core temperature in the two protocols. Sixty minutes after cooling commenced, the fetus was ventilated with oxygen (60-300 min), and the responses were followed for a 240-min period.

After completion of the birth-simulation study, the fetus and ewe were humanely killed with an overdose of pentobarbital sodium (Eutha-6 CII, Western Medical). The fetus was weighed to the nearest 10 g. The location of the thermistors was verified, as was the complete occlusion of the umbilical cord in protocol 1. The thermistors were then recalibrated to establish that their response characteristics had not changed during their implantation.

All these procedures were approved by the Animal Ethics Committee of Loma Linda University.

Analytical procedures

The results are given as means ± SE. Two-way analysis of variance with repeated measures was employed to determine the significance of change in response to various stimuli in each protocol. The significance of each added stimulus was tested post hoc by Fischer's least significant difference test for multiple comparisons between the value immediately before and at the end of the experimental period. Protocol 2, the birth simulation control study, was undertaken to assess the effect of theophylline by comparing the response to the adenosine antagonist administration to the response in the absence of the drug. The response was calculated by subtracting the mean of each animal in the initial 60 min of ventilation from the mean of each animal in the subsequent 90 min of adenosine antagonist administration (experimental) or continuing ventilation (control). The response mean and SE values for all animals in the experimental and control groups were then calculated. The significance of the response difference between the antagonist-treated and the untreated control groups was assessed with the t-test. P < 0.05 was considered significant.

RESULTS

Protocol 1: Administration of Theophylline During Birth Simulation

Table 1. Effect of administration of theophylline during birth simulation on fetal core temperature, arterial hemoglobin, pH and PaO2, and plasma glucose and theophylline levels Expt Period Fetal Core Temperature, °C Hemoglobin, g/dl Fetal Arterial Blood pH PaO2, Torr Glucose, mg/dl T,* mmol/l Control   0 min 39.6 ± 0.2  12.3 ± 0.7  7.34 ± 0.01  21.2 ± 2.5  23 ± 5  <5 Cool   +60 min 37.8 ± 0.1  13.5 ± 0.9  7.31 ± 0.02  14.9 ± 1.9  30 ± 8  <5 Ventilate   +120 min 37.8 ± 0.1  12.2 ± 0.8  7.26 ± 0.02  84.9 ± 28.6  43 ± 10  <5 Theophylline   +210 min 37.9 ± 0.2  12.3 ± 0.6  7.28 ± 0.02  124 ± 5  41 ± 6  150 ± 17  Occlude   +300 min 37.6 ± 0.3  12.7 ± 0.5  7.06 ± 0.13  110 ± 39  38 ± 3  423 ± 49 Values are means ± SE of 6 fetal sheep. PaO2, arterial PO2; T, plasma theophylline concentration.

To assess the well-being of the fetus before and during the simulation of birth, arterial hemoglobin, Pa_O2, and plasma glucose levels were monitored. All indexes came within the normal range during the control period. Fetal hemoglobin levels did not decrease significantly during the birth simulation despite repeated blood sampling. Upon ventilation, the fetal Pa_O2 rose 469% (P < 0.05) and remained elevated thereafter. Such levels of oxygenation are adequate in the newborn. Fetal glucose concentrations increased 43% during ventilation (P < 0.05) and remained high, indicating an abundance of available metabolic substrate during the simulation.

Control period. During the control period, the mean values for fetal core temperature, plasma FFA, and glycerol were within the normal range.

Cold exposure. During cooling, the fetal core temperature fell over a 60-min interval to 37.81 ± 0.13°C, averaging 1.78 ± 0.11°C less than the mean of the control period (P < 0.001). Plasma FFA (75.4 ± 6.7 µeq/l) and glycerol (151 ± 42 µmol/l) concentrations remained at low levels.

Ventilation with oxygen. During the subsequent ventilation with oxygen, theophylline administration, and umbilical cord occlusion phases, adjustment of the rate of cooling maintained the fetal core temperature at the level reached after 60 min of cooling. Upon elevation of arterial oxygen levels of >100 Torr, the fetal nonshivering thermogenic responses became apparent. Plasma levels of FFA and glycerol rose 176% to 208 ± 40 µeq/l (P < 0.05) and 87% to 282 ± 51 µmol/l (not significant), respectively, after 60 min of supplemental oxygen.

Theophylline administration. Within minutes of the theophylline infusion, an additional effect on fetal thermogenic responses became apparent. After 90 min of theophylline administration, plasma concentrations of FFA rose 99% (P < 0.01) to 415 ± 60 µeq/l, and glycerol levels rose 87% to 526 ± 135 µmol/l (P < 0.05).

Umbilical cord occlusion. Plasma FFA concentrations did not significantly change with umbilical cord occlusion. In contrast, cord occlusion increased glycerol concentrations 41% (P < 0.05) to 774 ± 203 µmol/l.

Protocol 2: Birth-Simulation Control

Table 2. Effect of control birth simulation on fetal core temperature, arterial hemoglobin, pH and PaO2, and plasma glucose levels Expt Period Fetal Core Temperature, °C Hemoglobin, g/dl Fetal Arterial Blood pH PaO2, Torr Glucose, mg/dl Control   0 min 39.8 ± 0.1  11.6 ± 0.5  7.34 ± 0.01  18.5 ± 1.2  19 ± 2  Cool   +60 min 38.2 ± 0.3  11.9 ± 0.6  7.34 ± 0.01  15.1 ± 1.8  25 ± 3  Ventilate   +120     min 38.0 ± 0.3  11.1 ± 0.5  7.31 ± 0.03  271 ± 60  33 ± 3    +210     min 37.9 ± 0.2  11.0 ± 0.5  7.33 ± 0.03  302 ± 67  41 ± 5    +300     min 37.9 ± 0.2  10.9 ± 0.7  7.34 ± 0.04  234 ± 71  42 ± 6 Values are means ± SE of 8 fetal sheep.

As in protocol 1, during the control period, all indexes of fetal well-being were within the normal range. Fetal hemoglobin concentrations remained stable until the end of the birth simulation. Upon ventilation, the Pa_O2 rose (P < 0.005) above the levels commonly seen in the newborn, and increased (P < 0.05) glucose concentrations ensured an adequate availability of metabolic substrate.

Control period. During the control period, the mean values for fetal core temperature, plasma FFA, and glycerol were within the normal range.

Cold exposure. During cooling, the fetal core temperature fell over a 60-min interval to 38.2 ± 0.3°C, averaging 1.65 ± 0.28°C less than the mean of the control period (P < 0.001). Plasma FFA (78.4 ± 16.5 µeq/l) and glycerol (83.7 ± 18.4 µmol/l) remained at low levels.

Ventilation with oxygen. During the subsequent ventilation with oxygen phase, the rate of cooling continued unchanged from that achieved during the cooling phase. When arterial oxygen levels rose above 200 Torr, the fetal nonshivering thermogenic responses became apparent. Plasma levels of FFA and glycerol rose 112% to 166 ± 34 µeq/l (P < 0.01) and 54% to 128 ± 32 µmol/l (not significant), respectively, after 60 min of supplemental oxygen. There was no further significant change in FFA concentrations. Glycerol levels, on the other hand, increased 62% (P < 0.05) after 150 min of supplemental oxygen to plateau at 208 ± 39 µmol/l. Comparison of the theophylline (protocol 1) administration response with the control (protocol 2) response reveals no difference in the fetal core temperature. Plasma FFA (P < 0.001) and glycerol (P < 0.05) concentrations were significantly greater during theophylline administration than in the corresponding control period.

DISCUSSION

Circulating plasma adenosine concentrations are three- to fivefold higher in the fetal sheep than in the adult (19). During the simulation of birth in utero, adenosine levels decline simultaneously with the initiation of nonshivering thermogenesis (19). Fetal cooling is followed by an increase in plasma adenosine and supplemental oxygenation is followed by a subsequent fall. Of most relevance are the responses to cord occlusion. At this time, plasma adenosine levels decline ~60% within 60 min (19), suggesting the placenta or umbilical vasculature as a likely source of a significant fraction of circulating adenosine. Similarly, in the human, adenosine has been reported to be released from isolated perfused cotyledons into the umbilical vein (24). In the fetus, the liver is another potential source of circulating adenosine (3), the contribution of which might change after cord occlusion with redirection of blood flow away from the placenta and liver; this remains to be tested experimentally. This series of studies tested the possible suppressive action of endogenous adenosine on ovine fetal nonshivering thermogenesis during the simulation of birth in utero.

Adenosine suppresses many metabolically dependent cellular activities by preventing the intracellular accumulation of adenosine 3,5-cyclic monophosphate (cAMP) (14). Direct evidence for this metabolic control has been found in the adult brain, heart, and white and brown adipose tissue of several species (14, 27). The regulatory role of adenosine as an inhibitor of - and -adrenergic-stimulated lipolysis in brown adipocytes is well documented (20, 29).

In brown adipose tissue, adenosine inhibits -adrenergic activation of adenyl cyclase activity (22) through its interaction with specific cell-surface A₁ receptors. The -adrenergic nonshivering thermogenic component is less responsive to the inhibitory actions of adenosine than is the -adrenergic component (21) and is not related to adenylate cyclase activity (23). This suggests the presence of a second mechanism, unrelated to adenylate cyclase, elicited by activation of adenosine receptors on brown adipose cells (21). Plasma levels of adenosine in the ovine fetus are well into the range that suppresses lipolysis in adult fat (27).

The administration of theophylline during protocol 1 was undertaken to investigate whether removal of the inhibitory effects of endogenous adenosine initiates nonshivering thermogenesis before umbilical cord occlusion in the fetal sheep. The infusion of theophylline began 90 min before cord occlusion and continued throughout the study. Theophylline significantly enhanced the nonshivering thermogenic responses before umbilical cord occlusion. The FFA and glycerol responses to oxygenation in the presence of theophylline were twofold greater than normal. Subsequent cord occlusion did not then alter FFA concentrations but did induce a significant rise in glycerol levels. The failure of plasma FFA levels to rise and the smaller increase of glycerol concentrations on umbilical cord occlusion in the presence of theophylline suggest that the removal of the inhibitory effects of endogenous adenosine before cord occlusion preempts the rise in the indexes of thermogenesis that occurs in the absence of theophylline.

Theophylline has been reported to increase cold resistance, possibly due to enhanced endogenous substrate mobilization (28). It may also do so through inhibition of adenosine 3,5-cyclic monophosphate phosphodiesterase and increasing intracellular cAMP concentrations (26). However, a 5 - 10% inhibition of phosphodiesterase requires a theophylline concentration of 560 µmol/l (18), so this action is unlikely in these studies, since the theophylline levels peaked at lower concentrations. Rather, it seems more likely that theophylline acted by blocking adenosine-mediated antilipolysis, an action that is fully exemplified at a theophylline level of <110 mmol/l and is particularly prominent under sympathetic-stimulated lipolysis (8). It has also been demonstrated that theophylline at low plasma concentrations can increase adrenal-medullary release of epinephrine and norepinephrine (10) and that theophylline may enhance peripheral sympathetic discharge of epinephrine by counteracting the inhibitory action of adenosine on such neural transmission (7). Furthermore, elevated plasma levels of free thyroxine and triiodothyronine have been reported after theophylline treatment (28). However, enhanced catecholamine and thyroid hormone levels induced by theophylline are likely to have only a minor effect on nonshivering thermogenesis before umbilical cord occlusion in the fetal sheep, in view of the minimal effect of administered norepinephrine (11) and triiodothyronine (17) in previous studies.

In summary, it may be concluded that there are increases in the indexes of nonshivering thermogenesis due to treatment with an adenosine-receptor antagonist. Furthermore, the inhibition of nonshivering thermogenesis by an adenosine analogue reported earlier suggests that cord occlusion may act via changes in circulating adenosine. Adenosine must thus be given serious consideration as one inhibitor of intrauterine thermogenesis. The further and continuing rise in plasma glycerol after cord occlusion, however, suggests that there is at least one other intrauterine thermogenic inhibitor physiologically linked to the onset of nonshivering thermogenesis at birth. Another intrauterine inhibitor candidate is an eicosanoid, and this possibility has been investigated elsewhere (9).

ACKNOWLEDGEMENTS

The skilled technical assistance of Shannon Bragg is gratefully acknowledged.

FOOTNOTES

Address for reprint requests: G. G. Power, Center for Perinatal Biology, School of Medicine, Loma Linda Univ., Loma Linda, CA 92350.

Received 19 May 1995; accepted in final form 6 August 1996.

REFERENCES

1.	Alexander, G., D. Nicol, and G. Thorburn. Thermogenesis in prematurely delivered lambs. In: Fetal and Neonatal Physiology. Proc. of the Sir Joseph Barcroft Centenary Symp., edited by K. S. Comline, K. W. Cross, and G. S. Dawes. Cambridge, UK: Cambridge Univ. Press, 1973, p. 410-417.
2.	Alexander, G., and D. Stevens. Sympathetic innervation and the development of structure and function of brown adipose tissue: studies on lambs chemically sympathectomized in utero with 6-hydroxydopamine. J. Dev. Physiol. 2: 119-137, 1980.
3.	Arnold, S. T., and R. L. Cysyk. Adenosine export from the liver: oxygen dependency. Am. J. Physiol. 251 (Gastrointest. Liver Physiol. 14): G34-G39, 1986.
4.	Ball, K. T., T. R. Gunn, G. G. Power, H. Asakura, and P. D. Gluckman. A potential role for adenosine in the inhibition of nonshivering thermogenesis in the fetal sheep. Pediatr. Res. 37: 303-309, 1995.
5.	Belle, H. van Uptake and deamination of adenosine by blood. Species differences, effect of pH, ions, temperature and metabolic inhibitors. Biochim. Biophys. Acta 192: 124-132, 1969.
6.	Falholt, K., B. Lund, and W. Falholt. An easy colorimetric micromethod for routine determination of free fatty acids in plasma. Clin. Chim. Acta 46: 105-111, 1973.
7.	Fredholm, B., and P. Hedqvist. Modulation of neurotransmission by purine nucleotides and nucleosides. Biochem. Pharmacol. 29: 1635-1643, 1980.
8.	Fredholm, B., and A. Sollevi. Regulation of lipolysis and circulation in adipose tissue. In: Purines: Pharmacology and Physiological Roles, edited by T. W. Stone. London: Macmillan, 1985, p. 223-232.
9.	Gunn, T. R., K. T. Ball, G. G. Power, and P. D. Gluckman. Withdrawal of placental prostaglandins permits thermogenic responses in fetal sheep brown adipose tissue. J. Appl. Physiol. 74: 998-1004, 1993.
10.	Higbee, M. D., M. Kumar, and S. P. Galant. Stimulation of endogenous catecholamine release by theophylline: a proposed additional mechanism of action for theophylline effects. J. Allergy Clin. Immunol. 70: 377-382, 1982.
11.	Hodgkin, D., R. D. Gilbert, and G. G. Power. In vivo brown fat response to hypothermia and norepinephrine in the ovine fetus. J. Dev. Physiol. 10: 383-391, 1988.
12.	Klein, A. H., A. Reviczky, P. Chou, J. Padbury, and D. A. Fisher. Development of brown adipose tissue thermogenesis in the ovine fetus and newborn. Endocrinology 112: 1662-1666, 1983.
13.	Mitchell, M. D., and A. P. F. Flint. Prostaglandin production by intra-uterine tissues from periparturient sheep: use of a superfusion technique. J. Endocrinol. 76: 111-121, 1978.
14.	Newby, A. C., Y. Worku, P. Meghji, M. Nakazawa, and A. Skladanowski. Adenosine: a retaliatory metabolite or not? News Physiol. Sci. 5: 67-70, 1990.
15.	Pinter, J. K., J. A. Hayashi, and J. A. Watson. Enzymatic assay of glycerol, dehydroxyacetone, and glyceraldehyde. Arch. Biochem. Biophys. 121: 404-414, 1967.
16.	Power, G. G., T. R. Gunn, B. M. Johnston, and P. D. Gluckman. Oxygen supply and the placenta limit thermogenic responses in fetal sheep. J. Appl. Physiol. 63: 1896-1901, 1987.
17.	Power, G. G., T. R. Gunn, B. M. Johnston, G. Nichols, and P. D. Gluckman. Umbilical cord occlusion but not increased plasma T3 or norepinephrine stimulate brown adipose tissue thermogenesis in the fetal sheep. J. Dev. Physiol. 11: 171-177, 1989.
18.	Rall, T. W. Evolution of the mechanism of action of methylxanthines: from calcium mobilizes to antagonists of adenosine receptors. Pharmacology 24: 277-287, 1982.
19.	Sawa, R., H. Asakura, and G. G. Power. Changes in plasma adenosine during simulated birth of fetal sheep. J. Appl. Physiol. 70: 1524-1528, 1991.
20.	Schimmel, R. J., M. E. Elliott, and V. C. Dehmel. Interactions between adenosine and 1-adrenergic agonists in regulation of respiration in hamster brown adipocytes. Mol. Pharmacol. 32: 26-33, 1987.
21.	Schimmel, R. J., and L. McCarthy. Effects of insulin, adenosine, and prostaglandin on -adrenergic-stimulated respiration in brown adipocytes. Am. J. Physiol. 250 (Cell Physiol. 19): C738-C743, 1986.
22.	Schimmel, R. J., L. McCarthy, and D. Dzierzanowski. Effects of pertussis toxin treatment on metabolism in hamster brown adipocytes. Am. J. Physiol. 249 (Cell Physiol. 18): C456-C463, 1985.
23.	Schimmel, R. J., L. McCarthy, and K. McMahon. 1-Adrenergic stimulation of hamster brown adipocyte respiration. Am. J. Physiol. 244: C362-C368, 1983.
24.	Slegel, P., H. Kitagawa, and M. H. Maguire. Determination of adenosine in fetal perfusates of human placental cotyledons using fluorescence derivatization and reversed-phase high-performance liquid chromatography. Anal. Biochem. 171: 124-134, 1988.
25.	Steinberg, D., M. Vaughan, P. J. Nestel, O. Strand, and S. Bergstrom. Effects of the prostaglandins on hormone-induced mobilization of free fatty acids. J. Clin. Invest. 43: 1533-1540, 1964.
26.	Swinyard, E. A. Respiratory drugs. In: Remington's Pharmaceutical Sciences, edited by A. Osol. Easton, PA: Mack, 1980, p. 809-814.
27.	Szillat, D., and L. Bukowiecki. Control of brown adipose tissue lipolysis and respiration by adenosine. Am. J. Physiol. 245 (Endocrinol. Metab. 8): E555-E559, 1983.
28.	Wang, L. C. H., S. F. P. Man, and A. N. Belcastro. Metabolic and hormonal responses in theophylline-increased cold resistance in males. J. Appl. Physiol. 63: 589-596, 1987.
29.	Woodward, J. A., and E. D. Saggerson. Effect of adenosine deaminase, N6-phenylisopropyladenosine and hypothyroidism on the responsiveness of rat brown adipocytes to norepinephrine. Biochem. J. 238: 395-403, 1986.