A study of the T system in rat heart.

The technique of extracellular space tracing with horseradish peroxidase is adapted for labeling the transverse tubular system (T system) in rat heart. In rat ventricular muscle the T system shows extensive branching and remarkable tortuosity. The T system can only be defined operationally, since it does not display specific morphological features throughout its entire structure. Owing to branching of the T system, a sizable proportion of the apposition between the T system and L system (or closed system) occurs at the level of longitudinal branches of the T system and is not restricted to the Z line region. The regions of apposition between the T system and L system are analyzed in rat ventricular muscle and skeletal muscle (diaphragm) and compared with the intercellular tight junctions (nexuses) of heart muscle by the use of a photometric method. The over-all thickness of the nexus is significantly smaller than that of T-L junctions in both cardiac and skeletal muscles. The thickness of the membranes of the T and L systems are not significantly different in the two muscles, but the gap between both membranes is larger in the heart. In atrial muscle the following two types of cells are found: (a) those cells with a well-developed T system in which the tubular diameter is quite uniform and the orientation predominantly longitudinal and, (b) cells with no T system, but with a well-developed L system. Atrial cells possessing a T system are richly provided with specific granules and show little micropinocytotic activity, whereas cells devoid of T system show intense micropinocytotic activity and few specific granules. The possible functional implications of these findings are discussed.

Crayfish muscle: permeability to sodium induced by calcium depletion.

Membrane of crayfish muscle fibers becomes selectively permeable to sodium when the calcium concentration of the bathing medium is reduced. Removal of calcium or its reduction below 1 or 2 millimole per liter causes large transient depolarizations up to 70 millivolts in amplitude. They resemble pro longed action potentials and occur only in the presence of sodium. The responses are abolished when tris(hydroxymethyl)aminomethane or lithium is substituted for sodium, and are blocked by tetrodotoxin even in the presence of sodium.

Neural regulation of insulin secretion in the dog.

The effects of stimulation of the mixed autonomic nerve to the dog pancreas has been studied under conditions in which both pancreaticoduodenal vein blood flow and insulin concentration were determined. Stimulation resulted in increased insulin output, which was blocked by prior administration of atropine. Blood flow was reduced by stimulation in proportion to the rate of stimulation. At 40 stimuli/s a maximum effect was found at 1 min with a gradual return toward base line despite continued application of the stimulus. Atropinization had no effect on blood flow changes. Insulin responses to 0.1 g/kg glucose were reduced on the average 40% by simultaneous stimulation of the pancreatic nerve at 40 cycles/s in atropinized animals. These studies establish this preparation as a reproducible model for the direct examination of autonomic influences on endocrine pancreatic function. From them it is concluded that the nerve supply to the endocrine pancreas of the dog is sufficient to inhibit insulin secretion by activation of the sympathetic nerves and to stimulate insulin secretion by activation of the parasympathetic nerves.

Extrapancreatic glucagon and glucagonlike immunoreactivity in depancreatized dogs. A quantitative assessment of secretion rates and anatomical delineation of sources.

The anatomical sites and the rates of extrapancreatic secretion of glucagon and of glucagon-like immunoreactivity (GLI) were assessed in dogs 2 h after pancreatectomy by catheterization of the gastrosplenic and mesenteric veins. Glucagon release from the gastrosplenic area approximated one-fourth that of a normal pancreas and rose from 0.25 to 1.0 ng/kg per min during arginine stimulation. Intestinal glucagon secretion was small and did not respond to arginine, suggesting that the stomach is the only important extrapancreatic source of glucagon. Glucagon concentrations attained by gastrosplenic secretion were in close proportion to those obtained during the administration of exogenous glucagon, indicating similar clearance rates of extrapancreatic and pancreatic glucagon, approximating 10 ml/kg per min.GLI secretion (0.3 ng eq/kg per min) was limited to the intestinal area and was transiently stimulated by arginine and exogenous glucagon. Base-line GLI clearance approximated 1 ml/kg per min. No insulin secretion could be detected. Gastrointestinal glucose uptake rose from 0.56 to 2.2 mg/kg per min after glucagon administration suggesting that as much as 10% of total glucose production can be taken up by the gastrointestinal tract. In two dogs both the stomach and pancreas were removed. Intestinal glucagon release remained small and did not increase during arginine administration. By contrast, GLI release was stimulated by both arginine and exogenous glucagon.

Glucagon release induced by pancreatic nerve stimulation in the dog.

A direct neural role in the regulation of immunoreactive glucagon (IRG) secretion has been investigated during stimulation of mixed autonomic nerves to the pancreas in anesthetized dogs. The responses were evaluated by measurement of blood flow and hormone concentration in the venous effluent from the stimulated region of pancreas. Electrical stimulation of the distal end of the discrete bundles of nerve fibers isolated along the superior pancreaticoduodenal artery was invariably followed by an increase in IRG output. With 10-min periods of nerve stimulation, the integrated response showed that the higher the control glucagon output, the greater was the increment. Atropinization did not influence the response to stimulation. That the preparation behaved in physiologic fashion was confirmed by a fall in IRG output, and a rise in immunoreactive insulin (IRI) output, during hyperglycemia induced by intravenous glucose (0.1 g/kg). The kinetics of this glucose effect on IRG showed characteristics opposite to those of nerve stimulation: the lower the control output, the less the decrement. Furthermore, during the control steady state, blood glucose concentration was tightly correlated with the IRI/IRG molar output ratio, the function relating the two parameters being markedly nonlinear. Injection or primed infusion of glucose diminished the IRG response to simultaneous nerve stimulation. Measurement of IRG was inferred to reflect response of pancreatic glucagon secretion on the basis of the site of sample collection (the superior pancreaticoduodenal vein), the absence of changes in arterial IRG, and similar responses being obtained using an antibody specific for pancreatic glucagon. THESE STUDIES SUPPORT A ROLE FOR THE AUTONOMIC NERVOUS SYSTEM IN THE CONTROL OF GLUCAGON SECRETION: direct nerve stimulation induces glucagon release. Such sympathetic activation may be interpreted as capable of shifting the sensitivity of the A cell to glucose in the direction of higher glycemia for a given glucagon output. The experimental model employed is valid for further studies of regulatory mechanisms of endocrine pancreatic function in vivo.

Membrane potential of brown adipose tissue. A suggested mechanism for the regulation of thermogenesis.

Membrane potentials were recorded in isolated segments of interscapular brown adipose tissue from rats. After equilibration at 29 degrees C in Krebs-Ringer bicarbonate buffer a mean value of -51 +/- 4 mv (SD) was found. This level could be maintained for up to 5 hr. The mean effective membrane resistance was 1.35 +/- 0.45 megohm. The membrane potential was a function of the extracellular potassium concentration. Ouabain (10(-6)-10(-3)M) and incubation in K-free buffer produced progressive depolarization. Epinephrine and norepinephrine in concentrations as low as 10(-8) g/ml produced a prompt depolarization. Cooling of the tissue and lowering of the oxygen tension caused a marked and reversible decrease in the membrane potential. In tissue obtained from cold-adapted rats, the membrane potential was considerably diminished. 6Assuming that the membrane potential is some function of the Na permeability of the plasma membrane it is suggested that an increase in the rate of active Na-K transport and ensuing ADP formation might contribute to the increase in respiration seen during exposure to thermogenic stimuli.

Alpha- and beta-adrenergic mediation of changes in metabolism and Na/K exchange in rat brown fat.

Double- and triple-barreled ion-sensitive microelectrodes were used to measure changes in extracellular K+ and Na+ concentrations ([K+]o, [Na+]o) in brown fat. Redox states of different respiratory enzymes were measured simultaneously in order to correlate ion movements with metabolic activity. Trains of stimuli applied to the efferent nerves evoked two distinct increases in [K+]o. A first, small, rapid increase occurred within 10 s and accompanied a first, rapid membrane depolarization. A second, slow increase of [K+]o occurred several minutes after stimulation and accompanied a second, slow depolarization. A few seconds after stimulation onset, while the membrane was repolarizing and shifts in redox states indicated increases in lipolysis and respiration, [K+]o decreased. The [K+]o decrease was accompanied by an increase in [Na+]o, and could be partly blocked by ouabain. Phentolamine, an alpha-antagonist that blocks the first depolarization, also blocked the first, rapid [K+]o increase and part of the subsequent decrease. Propranolol, a beta-antagonist, had little effect on the first depolarization and the first increase in [K+]o, but blocked part of the subsequent [K+]o decrease and the second, slow [K+]o increase. The changes in [K+]o were almost completely abolished in the presence of both antagonists. It is concluded that brown adipocytes take up K+ and simultaneously lose Na+ in response to the interaction of noradrenaline with alpha- and beta-receptors, and this indicates a very early stimulation of the Na+ pump.

Effects of caffeine on crayfish muscle fibers. II. Refractoriness and factors influencing recovery (repriming) of contractile responses.

When caffeine evokes a contraction, and only then, crayfish muscle fibers become refractory to a second challenge with caffeine for up to 20 min in the standard saline (5 mM K(o)). However, the fibers still respond with contraction to an increase in K(o), though with diminished tension. Addition of Mn slows recovery, but the latter is greatly accelerated during exposure of the fiber to high K(o), or after a brief challenge with high K(o). Neither the depolarization induced by the K, nor the repolarization after its removal accounts for the acceleration, which occurs only if the challenge with K had itself activated the contractile system; acceleration is blocked when contractile responses to K are blocked by reducing the Ca in the bath or by adding Mn. Recovery is accelerated by redistribution of intracellular Cl and by trains of intracellularly applied depolarizing pulses, but not by hyperpolarization. The findings indicate that two sources of Ca can be mobilized to activate the contractile system. Caffeine mobilizes principally the Ca store of the SR. Depolarizations that are induced by high K(o), by transient efflux of Cl, or by intracellularly applied currents mobilize another source of Ca which is strongly dependent upon the entry of Ca from the bathing medium. The sequestering mechanism of the SR apparently can utilize this second source of Ca to replenish its own store so as to accelerate recovery of responsiveness to a new challenge with caffeine.

Adrenoceptor heterogeneity in human white adipocytes differentiated in culture as assessed by cytosolic free calcium measurements.

Changes in intracellular calcium concentration [Ca2+]i in response to norepinephrine (NE) and to various adrenergic agonists were monitored by dual excitation microfluorimetry in single human adipocytes differentiated in culture and loaded with fura-2 acetoxymethyl ester (fura-2 AM). The addition of NE elicited increases in [Ca2+]i that were depending on the cell, (1) either rapid (time to peak: 9 +/- 3 s), large, and transient; or (2) slow (time to peak: 125 +/- 8 s), small, and sustained. The rapid and large [Ca+]i response, which was inhibited by 90% by the alpha 1-antagonist prazosin and only by 20% by the non-specific beta antagonist (-)-propranolol, was considered to be mediated by the alpha 1-adrenoceptor. In fact, an alpha 1A-adrenoceptor was found to be expressed in human white adipose tissue. Consecutive additions of beta-agonists specific for each subtype of alpha-adrenoceptor enabled the characterization of four cell populations with different response patterns: 47% of the cells had alpha 1- and beta 1-, beta 2- and beta 3-induced [Ca2+]i responses; 29% had only beta 1-, beta 2-, beta 3-responses; 14% had alpha 1- and beta 3-responses, and 10% had only an alpha 1-response. Taken together, these results show that in differentiated human adipocytes: (1) alpha 1- and beta-adrenergic stimulations induce [Ca2+]i increases with different kinetics and amplitudes; (2) there is a beta 3-adrenergic response similar to the beta 1- or beta 2-adrenergic responses; and (3) there is a marked adrenoceptor heterogeneity.

Is there a sympathetic regulation of the efficiency of energy utilization?

Brown adipose tissue, a well known effector of regulatory thermogenesis found in mammals, is unique in its ability to steadily increase its heat production several fold for very long periods of time. It constitutes a shunt of energy flow between food intake and heat dissipation, it is activated through its sympathetic nerve supply. There are evidence in the rat, that brown adipose tissue is activated following overfeading, thus decreasing food efficiency and determining resistance to obesity. Genetically obese (ob/ob) mice fed and kept at 22 °C lack the possibility of activating their brown fat energy shunt; they are known to be poorly resistant to cold stress despite their large insulation. This is taken as a further circumstantial evidence of an overlap in thermal and food efficiency regulatory systems in rodents through sympathetically controlled brown fat as a common effector.

Alterations of brown adipose tissue in genetically obese (ob/ob) mice. I. Demonstration of loss of metabolic response to nerve stimulation and catecholamines and its partial recovery after fasting or cold adaptation.

Metabolic responses to electrical nerve stimulation or norepinephrine of isolated interscapular brown adipose tissue (BAT) from lean and ob/ob mice were studied using either continuous monitoring of the NAD(P)H/NAD(P) redox state or direct microcalorimetry. The responses to these sympathetic stimuli were not significantly different from zero in BAT of ob/ob mice kept at 22 C and fed ad libitum. The metabolic rate of BAT of lean mice was stimulated 3-fold by norepinephrine. ob/ob mice are hyperglycemic and hyperinsulinemic; cold adaptation further increased their plasma glucose, and fasting decreased the levels of both glucose and insulin to normal values. Both fasting and cold adaptation at 5 C partially restored the tissue metabolic response of ob/ob mice, whereas a decreased sensitivity was observed in the tissue of lean mice. The results of these experiments are compatible with the hypothesis that the impaired capacity of BAT of ob/ob mice to produce heat could be one of the causes of their high food efficiency and their inability to withstand acute cold exposure.

Increased plasma clearance rate of thyroxine despite decreased 5'-monodeiodination: study with a peroxisome proliferator in the rat.

In euthyroid rats a 17-day treatment with nafenopin, a hypolipidemic agent and peroxisome proliferator, decreased serum total and free T4 concentrations to 32 +/- 5% and 62 +/- 8% (mean +/- SEM; n = 10), respectively, with no change in serum T3 and TSH concentrations. In methimazole-treated rats infused with 3 nmol T4/day/100 g BW, the nafenopin inhibitory effect was not significantly different from that in euthyroid rats. Nafenopin treatment had the following effects on peripheral T4 and T3 metabolism in euthyroid rats. The plasma clearance rate of T4 (PCR), which was measured by Alzet minipump infusion of tracer, was increased 2-fold (1.58 +/- 0.09 vs. 0.82 +/- 0.06 ml/h.100 g BW; P less than 0.001; n = 5), while the PCR of T3 was decreased (37.5 +/- 1.3 vs. 53.8 +/- 1.8; P less than 0.001; n = 5). The fecal clearance rate of radioactivity derived from T4 was increased 2-fold (1.93 +/- 0.10 vs. 0.77 +/- 0.07 ml/h.100 g BW), whereas the urinary clearance rate was not significantly modified. The 5'-deiodinase (5'D) activity, measured by deiodination of labeled rT3, was strongly inhibited in liver and kidney, not modified in brown fat and anterior pituitary, and increased in cerebral cortex. In methimazole-treated rats substituted with isopropyl-diiodothyronine only hepatic 5'D activity was decreased. It is concluded that the decrease in serum total and free T4, without alteration in serum T3 and TSH concentrations, resulting from nafenopin treatment is mainly due to changes in peripheral T4 and T3 metabolism, since it is also observed in T4-substituted animals. The increased PCR of T4 cannot be explained by an increase in deiodination activity, since the major 5'D pathways are inhibited after nafenopin treatment, and the urinary clearance rate is not modified. It can partly be explained by an increase in the fecal clearance rate of T4, which could be due to an increase in glucoronoconjugation. In addition, nafenopin was found to be a weak competitor of T4 binding to serum proteins, leading to a small increase in the free T4 fraction which might also contribute to the increased T4 PCR. The decrease in T3 PCR remains to be explained.

Alterations of brown adipose tissue in genetically obese (ob/ob) mice. II. Studies of beta-adrenergic receptors and fatty acid degradation.

beta-Receptor number, norepinephrine-stimulated adenylate cyclase activity and lipolysis, octanoate-induced NAD(P) redox changes, and heat production were studied in brown adipose tissue (BAT) of lean and obese (ob/ob) mice. beta-Receptor number was increased 1.54-fold in purified brown adipocyte plasma membrane of ob/ob mice compared to that in lean controls. This increase was reversed by cold adaptation (5 C). Basal and norepinephrine-stimulated adenylate cyclase values were not different in the two groups. Norepinephrine stimulated lipolysis at 10 nM in BAT of lean mice, but only at 10 microM in BAT ob/ob mice. Octanoate produced an increase in the NAD(P) redox state in BAT of lean mice, but it did not modify the NAD(P) redox state in BAT of ob/ob mice. Concomitantly, octanoate increased heat production 3-fold in BAT of lean mice, but did not promote any significant increase in heat production in BAT of ob/ob mice. These two parameters were restored toward values observed in lean mice when the ob/ob mice were adapted to a cold environment. The data indicate that BAT of ob/ob mice exhibits three alterations; one at the level of the beta-receptor, one at the level of the lipolytic response to norepinephrine, and one at the level of fatty acid activation and/or beta-oxidation.

Brown adipose tissue metabolism in streptozotocin-diabetic rats.

Defects of both diet-induced thermogenesis and cold tolerance have been reported for streptozotocin-diabetic rats. Since brown adipose tissue (BAT) is a major effector of both diet- and cold-induced thermogenesis in the rat, the possible cause of these defects was investigated by comparing BAT metabolism under basal conditions and during activation by nerve stimulation, norepinephrine (NE), or octanoate addition in both streptozotocin-diabetic rats and in controls. The following metabolic indices were measured in rat interscapular BAT (IBAT): 1) tissue composition, 2) heat production rate as measured by direct microcalorimetry, 3) redox state of flavoproteins linked to the acyl-coenzyme A dehydrogenase pathway as measured by reflection spectrometry, 4) redox state of NAD(P) as measured by surface-emitted fluorescence, and 5) fatty acid activation and beta-oxidation activities in IBAT homogenate. In streptozotocin-diabetic rats, IBAT was atrophied (DNA content unmodified, protein and lipid content decreased). The basal and NE-stimulated total heat production rates showed a 75% and 56% decrease, respectively. The specific activity of fatty acid beta-oxidation as measured by flavoprotein redox state or enzymatically was decreased by 52% and 59%, respectively. The basal redox level of NAD(P) was about 3 times higher than in the controls and NE stimulation resulted in oxidation in contrast to the reduction observed in control tissues. These results show that the metabolic capacity of IBAT from streptozotocin-diabetic rats is decreased and further suggest that the reduced capacity for beta-oxidation contributes significantly to the metabolic alteration.

Modulation of beta-oxidation and proton conductance pathway of brown adipose tissue in hypo- and hyperinsulinemic states.

The metabolic capacity of interscapular brown adipose tissue of hypoinsulinemic (diabetic) rats is decreased and a reduced beta-oxidative capacity contributes to this metabolic alteration. It was thus of interest to compare, in diabetic and in chronically (8 days) insulin-infused rats, the beta-oxidative capacity and indices of the thermogenic state (GDP-binding and 32 000 Mr protein) in this tissue. Mitochondrial GDP-binding and 32 000 Mr protein were both decreased in diabetic rats compared to appropriate controls and markedly increased as was also the beta-oxidative capacity in hyperinsulinemic rats.

Adaptive role of energy expenditure in modulating body fat and protein deposition during catch-up growth after early undernutrition.

After 10 or 30 d of growth arrest due to undernutrition during the early weaning period, rehabilitated rats fed ad libitum showed the phenomenon of catch-up growth, ie, gains in body weight and body protein were greater by 50% and 25%, respectively, (P < 0.001) than those of controls with similar starting body weight [ie, weight-matched (WM) controls]. These increases, however, were entirely dependent on the higher food intake because they failed to occur when food intakes were maintained at similar amounts to those for WM controls. In contrast, independent of whether the rehabilitated groups were spontaneously hyperphagic relative to WM controls or made normophagic by pair-feeding to WM controls, the gross energetic efficiency was increased (P < 0.01) and body fat gain was more elevated (2-2.5 fold, P < 0.001) during the first 2 wk of refeeding compared with WM controls--differences that were uninfluenced by the duration of growth arrest. Taken together, these studies suggest that the often reported impressive gains in body fat during recovery from malnutrition may result not only from unbalanced diets or excess dietary intake, but also from a transitory enhancement in the efficiency of food utilization and a shift in energy partitioning in favor of an acceleration for the replenishment of fat stores.

Adaptive changes in energy expenditure during refeeding following low-calorie intake: evidence for a specific metabolic component favoring fat storage.

Studies in this report investigate the existence and quantitative contribution of a metabolic component to an apparently elevated efficiency of energy utilization during refeeding following low-calorie consumption. Energy balance during a 25-d refeeding period was assessed in rats that had been food restricted (at 50% of normal intake) for either 30 or 10 d. Relative to weight-matched controls (with comparable lean tissue mass and similar food intake), refeeding following both periods of low food consumption (30 or 10 d) was associated with a 10% reduction in energy expenditure (p less than 0.001). Analysis of body composition revealed that virtually all the energy saved as a result of this metabolic (as opposed to tissue mass) adaptation was deposited as fat rather than protein. After the energy cost for depositing the extra fat is accounted for, the metabolic component represents a net 15% lower energy expenditure when normal food intake is resumed. These adaptive changes in energy expenditure may thus constitute an important mechanism for the rapid replenishment of energy stores in preparation for recurring food shortages and may also underlie the ease with which the obese condition is rapidly reachieved after weight loss.

Poststarvation hyperphagia and body fat overshooting in humans: a role for feedback signals from lean and fat tissues.

An increase in the sensation of hunger and overeating after a period of chronic energy deprivation can be part of an autoregulatory phenomenon attempting to restore body weight. To gain insights into the role of fat and lean tissue depletion as determinants of such a hyperphagic response in humans, we reanalyzed the individual data on food intake and body composition available for the 12 starved and refed men in the classical Minnesota Experiment after a shift from a 12-wk period of restricted refeeding to an ad libitum refeeding period of 8 wk. For each individual, the following were determined: 1) the total hyperphagic response during the ad libitum refeeding period, calculated as the energy intake in excess of that during the prestarvation (control) period; 2) the degree of fat recovery and that of fat-free-mass (FFM) recovery before ad libitum refeeding, calculated as the deviation in fat and FFM from their respective prestarvation values (ie, the amount of fat or FFM before ad libitum refeeding as a percentage of fat or FFM during the control period); and 3) the deficit in energy intake before ad libitum refeeding, calculated as the difference between the energy intake during the period of restricted refeeding and that during the control period. The results indicate that 1) the total hyperphagic response is inversely correlated with the degree of fat recovery (r = -0.6) as well as with that of FFM recovery (r = -0.5), 2) the correlation between hyperphagia and FFM recovery persists after adjustment for fat recovery, and 3) the correlations between hyperphagia and fat recovery or FFM recovery persist after adjustment for the variance in the energy deficit during the preceding period of restricted refeeding. Taken together, these results in humans suggest that poststarvation hyperphagia is determined to a large extent by autoregulatory feedback mechanisms from both fat and lean tissues. These findings, which have implications for both the treatment of obesity and for nutritional rehabilitation after malnutrition and cachexia, have been integrated into a compartmental model of autoregulation of body composition, and can be used to explain the phenomenon of poststarvation overshoot in body fat.

Efficacy of a green tea extract rich in catechin polyphenols and caffeine in increasing 24-h energy expenditure and fat oxidation in humans.

BACKGROUND: Current interest in the role of functional foods in weight control has focused on plant ingredients capable of interfering with the sympathoadrenal system. OBJECTIVE: We investigated whether a green tea extract, by virtue of its high content of caffeine and catechin polyphenols, could increase 24-h energy expenditure (EE) and fat oxidation in humans. DESIGN: Twenty-four-hour EE, the respiratory quotient (RQ), and the urinary excretion of nitrogen and catecholamines were measured in a respiratory chamber in 10 healthy men. On 3 separate occasions, subjects were randomly assigned among 3 treatments: green tea extract (50 mg caffeine and 90 mg epigallocatechin gallate), caffeine (50 mg), and placebo, which they ingested at breakfast, lunch, and dinner. RESULTS: Relative to placebo, treatment with the green tea extract resulted in a significant increase in 24-h EE (4%; P < 0.01) and a significant decrease in 24-h RQ (from 0.88 to 0.85; P < 0.001) without any change in urinary nitrogen. Twenty-four-hour urinary norepinephrine excretion was higher during treatment with the green tea extract than with the placebo (40%, P < 0.05). Treatment with caffeine in amounts equivalent to those found in the green tea extract had no effect on EE and RQ nor on urinary nitrogen or catecholamines. CONCLUSIONS: Green tea has thermogenic properties and promotes fat oxidation beyond that explained by its caffeine content per se. The green tea extract may play a role in the control of body composition via sympathetic activation of thermogenesis, fat oxidation, or both.

Dietary supplementation with fish oil enhances in vivo synthesis of tumor necrosis factor.

Studies reported here investigate the influence of dietary fat types on cytokine production in response to endotoxin (LPS) challenge. Tumor necrosis factor (TNF) serum levels were markedly higher (by 10-fold) in mice fed chronically a diet rich in fish oil rather than either a diet rich in corn or coconut oil or a low fat diet. This in vivo hyper-responsiveness in LPS-induced TNF production following fish oil consumption concorded with similar exaggerated in vitro TNF release from macrophages exposed to LPS. These data suggest that high consumption of fish oils, by virtue of their high content of omega-3 polyunsaturated fatty acids, can lead to an exaggerated production of mediators of inflammation with potentially adverse consequences on the outcome and severity of infectious diseases.