Characterization of the inhibitor sensitivity of the coenzyme A transport system in isolated rat heart mitochondria.

The effect of protein labeling agents on coenzyme A (CoA) transport into isolated rat heart mitochondria was studied. CoA transport was substantially inhibited by sulfhydryl reagents (mersalyl, pCMB) as well as by the tyrosine-selective reagent N-acetylimidazole. The effect of pCMB was reversed by DTT. Moreover, CoA uptake was completely abolished by agents selective for lysine and amino terminal residues (pyridoxal 5-phosphate, dansyl chloride). In contrast arginine-selective reagents (2, 3-butanedione, phenylglyoxal) caused considerably less inhibition of CoA uptake. Moreover, partial inhibition of transport was observed with the stilbene disulfonic acid derivatives DIDS and SITS. Finally, measurement of the effects of the labeling agents on the mitochondrial membrane potential indicated that the inhibition of CoA transport into mitochondria is not a secondary effect that arises from an alteration in the electric potential gradient across the inner mitochondrial membrane. These results provide the first information on the types of amino acid residues that may be essential to the CoA transport mechanism and provide additional support for the existence of a CoA transport protein within the mitochondrial inner membrane. Furthermore, the identification of effective inhibitors of the CoA transport system will greatly facilitate the functional reconstitution of this transporter in a proteoliposomal system following its solubilization and purification.

Cardiovascular abnormalities associated with human and rodent obesity.

Obesity is a major risk factor for cardiovascular disease. However, a direct link between these two states is difficult to establish, since obesity frequently occurs with other disease states such as diabetes, hypertension and atherosclerosis. Clinical studies have clearly shown that uncorrected obesity is associated with cardiac hypertrophy and compromised ventricular function. A number of rodent models of obesity have been studied in terms of cardiovascular adaptations. Cardiac function of the obese Zucker rat appears to be normal at a younger age. Only after several months is depression in cardiac function discernable. These animals are mildly hypertensive, but do not exhibit the characteristic increase in cardiac output associated with human obesity. A unique characteristic of JCR:LA-cp rat is that they develop atherosclerotic and myocardial lesions. Hearts from these animals will maintain normal function when perfused with physiological levels of calcium. At higher calcium concentrations, however, mechanical function becomes impaired. Dietary-induced obese rats exhibit many of the hemodynamic alterations associated with human obesity, but there is no evidence to-date that these animals will develop severe cardiac depression. Short-term weight reduction apparently has beneficial cardiovascular effects, but weight cycling may be harmful. Given the widespread occurrence of obesity, further studies are warranted to characterize the cardiac manifestations of this condition.

Evidence for net uptake and efflux of mitochondrial coenzyme A.

Coenzyme A transport was studied by determining [14C]CoA associated with isolated rat heart mitochondria. HPLC analysis of a mitochondrial extract obtained following incubation with [14C] CoA revealed an increase in [14C] CoA. In the presence of pyruvate or alpha-ketoglutarate, [14C]CoA associated with mitochondria was converted to acetyl- or succinyl-[14C]CoA, respectively, demonstrating the intramitochondrial localization of transported CoA. Net uptake of CoA was demonstrated by the findings that the increase in mitochondrial content of CoA following incubation with CoA was equal to the values of CoA uptake obtained from experiments using [14C]CoA. Sequestration of intramitochondrial CoA as metabolically inert derivatives with maleate stimulated CoA uptake, supporting the concept of unidirectional CoA uptake rather than exchange. Altering the membrane electrochemical gradient with valinomycin, nigericin, calcium, phosphate or a combination of phosphate and calcium caused efflux of endogenous CoA. The largest efflux was observed with valinomycin or a combination of Ca2+ and Pi. The Ca2+ and Pi-induced CoA efflux was effectively prevented by succinate or pyruvate. The results suggest that the uptake process, which is dependent on the membrane electrical gradient can be reversed by dissipating the electrical gradient. The relevance of CoA efflux induced by Ca2+ and Pi is discussed with respect to reperfusion injury following ischemic damage. Other factors regulating the maintenance of CoA within the mitochondrial matrix include the matrix pH and the acylation state of CoA.

Pantothenic acid in health and disease.

In summary, the vitamin pantothenic acid is an integral part of the acylation carriers, CoA and acyl carrier protein (ACP). The vitamin is readily available from diverse dietary sources, a fact which is underscored by the difficulty encountered in attempting to induce pantothenate deficiency. Although pantothenic acid deficiency has not been linked with any particular disease, deficiency of the vitamin results in generalized malaise clinically. In view of the fact that pantothenate is required for the synthesis of CoA, it is surprising that tissue CoA levels are not altered in pantothenate deficiency. This suggests that the cell is equipped to conserve its pantothenate content, possibly by a recycling mechanism for utilizing pantothenate obtained from degradation of pantothenate-containing molecules. Although the steps involved in the conversion of pantothenate to CoA have been characterized, much remains to be done to understand the regulation of CoA synthesis. In particular, in view of what is known about the in vitro regulation of pantothenate kinase, it is surprising that the enzyme is active in vivo, since factors that are known to inhibit the enzyme are present in excess of the concentrations known to inhibit the enzyme. Thus, other physiological regulatory factors (which are largely unknown) must counteract the effects of these inhibitors, since the pantothenate-to-CoA conversion is operative in vivo. Another step in the biosynthetic pathway that may be rate limiting is the conversion of 4'-phosphopantetheine (4'-PP) to dephospho-CoA, a step catalyzed by 4'-phosphopantetheine adenylyl-transferase. In mammalian systems, this step may occur in the mitochondria or in the cytosol. The teleological significance of these two pathways remains to be established, particularly since mitochondria are capable of transporting CoA from the cytosol. Altered homeostasis of CoA has been observed in diverse disease states including starvation, diabetes, alcoholism, Reye syndrome (RS), medium-chain acyl CoA dehydrogenase deficiency, vitamin B12 deficiency, and certain tumors. Hormones, such as glucocorticoids, insulin, and glucagon, as well as drugs, such as clofibrate, also affect tissue CoA levels. It is not known whether the abnormal metabolism observed in these conditions is the result of altered CoA metabolism or whether CoA levels change in response to hormonal or nonhormonal perturbations brought about in these conditions. In other words, a cause-effect relation remains to be elucidated. It is also not known whether the altered CoA metabolism (be it cause or result of abnormal metabolism) can be implicated in the manifestations of a disease. Besides CoA, pantothenic acid is also an integral part of the ACP molecule.(ABSTRACT TRUNCATED AT 400 WORDS)

Coenzyme A metabolism in vitamin B-12-deficient rats.

Vitamin B-12 (cobalamin) deficiency results in decreased L-methylmalonyl-coenzyme A (CoA) mutase activity. The consequence of this defect on the cellular CoA pool was studied in rats with functional vitamin B-12 deficiency induced by administration of the cobalamin analogue hydroxy-cobalamin [c-lactam] or by dietary vitamin B-12 deficiency. Both types of vitamin B-12 deficiency were associated with methylmalonic acidemia (100-300-fold increases in plasma methylmalonic acid concentration compared with controls), but overall fuel homeostasis was intact. Liver from rats treated with hydroxy-cobalamin [c-lactam] contained a threefold greater concentration of total CoA (free CoA plus all acyl-CoA) compared with saline-treated rats. Fractionation of the CoA pool revealed higher levels of CoA, propionyl-CoA, methyl-malonyl-CoA, acid-insoluble CoA, as well as total CoA in the rats treated with hydroxy-cobalamin [c-lactam] compared with controls. Similar increases in liver CoA content were seen in dietary vitamin B-12 deficiency in both the fed and fasted states. To examine the hypothesis that sequestration of hepatic CoA as propionyl-CoA and methylmalonyl-CoA could increase CoA biosynthesis, the effect of propionate on CoA biosynthesis was studied in hepatocytes isolated from control rats. Propionate (1 mM) increased the formation of 14C-CoA from [14C]pantothenate (10 microM) by 27% in the hepatocyte system. When butyrate (1 mM) was provided as substrate, propionate (10 mM) increased [14C]CoA formation by 63%.(ABSTRACT TRUNCATED AT 250 WORDS)

Dependence of mitochondrial coenzyme A uptake on the membrane electrical gradient.

Coenzyme A (CoA) transport was studied in isolated rat heart mitochondria. Uptake of CoA was assayed by determining [3H]CoA associated with mitochondria under various conditions. Various oxidizable substrates including alpha-ketoglutarate, succinate, or malate stimulated CoA uptake. The membrane proton (delta pH) and electrical (delta psi) gradients, which dissipated with time in the absence of substrate, were maintained at their initial levels throughout the incubation in the presence of substrate. Addition of phosphate caused a concentration-dependent decrease of both delta pH and CoA uptake. Nigericin also dissipated the proton gradient and prevented CoA uptake. Valinomycin also prevented CoA uptake into mitochondria. Although the proton gradient was unaffected, the electrical gradient was completely abolished in the presence of valinomycin. Addition of 5 mM phosphate 10 min after the start of incubation prevented further uptake of CoA into mitochondria. A rapid dissipation of the proton gradient upon addition of phosphate was observed. Addition of nigericin or valinomycin 10 min after the start of incubation also resulted in no further uptake of CoA into with mitochondria; valinomycin caused an apparent efflux of CoA from mitochondria. Uptake was found to be sensitive to external pH displaying a pH optimum at pHext 8.0. Although nigericin significantly inhibited CoA uptake over the pHext range of 6.75-8, maximal transport was observed around pHext 8.0-8.25. Valinomycin, on the other hand, abolished transport over the entire pH range. The results suggest that mitochondrial CoA transport is determined by the membrane electrical gradient. The apparent dependence of CoA uptake on an intact membrane pH gradient is probably the result of modulation of CoA transport by matrix pH.

Mitochondrial synthesis of coenzyme A is on the external surface.

Synthesis of coenzyme A (CoA) in isolated rat heart mitochondria was studied. Mitochondria synthesized CoA from 4'phosphopantetheine (4'PP), a precursor of CoA which is synthesized from pantothenic acid in the cytosol. The synthesis was time dependent and was absolutely dependent on the presence of external ATP under the conditions used. The rate of synthesis was increased either by increasing the pH from 7.4 to 8.5 or by adding 0.01% deoxycholate to the incubation medium. To determine whether the synthesis was intra- or extramitochondrial, mitochondria were separated from the incubation mixture by centrifugation and assayed for CoA. The amount of CoA appearing in the supernatant after 30 mins of incubation increased with increasing concentrations of 4'PP while that in the mitochondrial pellet remained constant over the concentration range studied. Synthesis of CoA from 4'PP was not affected by the uncoupler 2,4-dinitrophenol, or by carboxyatractyloside, or by a combination of the two drugs. The combination was used in an effort to deplete intramitochondrial ATP and to prevent external ATP from entering the mitochondria, thus resulting in mitochondria devoid of matrix ATP. The absolute dependence of synthesis on external ATP, the appearance of newly synthesized CoA in the incubation buffer and the ability of mitochondria lacking matrix-ATP to synthesize CoA suggest that the mitochondrial enzymes responsible for synthesis of CoA from 4'PP are on the outer membrane or on the outer side of the inner membrane.

A transport system for coenzyme A in isolated rat heart mitochondria.

The ability of isolated rat heart mitochondria to take up coenzyme A (CoA) from the incubation medium was studied. Mitochondria accumulated CoA in a time- and concentration-dependent manner. The accumulation process occurred in two phases. Within the first 30 s of incubation, mitochondrial content of CoA increased, and this phase did not plateau in the concentration range studied. Following this initial increase, a second slower phase of CoA accumulation occurred which plateaued around 50 microM CoA. The initial phase was decreased significantly by ATP or by carboxyatractyloside. In contrast, the presence of ATP or carboxyatractyloside did not affect the second phase. Decreasing the temperature from 30 to 0 degrees C did not affect the initial phase, but the second phase was almost abolished. In the presence of metabolic inhibitors (either 2,4-dinitrophenol or a combination of rotenone and antimycin), the initial "binding" phase was not affected; but the second "uptake" phase was abolished. These results suggest that the first phase of mitochondrial CoA accumulation is probably CoA binding to adenine recognizing sites on the mitochondria while the second phase may represent a specific uptake process for CoA which, although not directly ATP-dependent, is energy-dependent.

Effects of triiodothyronine and carnitine therapy on myocardial dysfunction in diabetic rats.

Streptozocin-diabetic rats were treated with a combination of triiodothyronine and carnitine for 6 weeks. These compounds were used as they are known to correct the diabetes-induced depression of cardiac myosin ATPase and sarcoplasmic reticular (SR) calcium uptake, respectively. Myocardial performance, which was assessed using the working heart preparation, revealed a depression of function in untreated diabetics when compared with controls at most left atrial filling pressures. Hearts from diabetic rats treated with the combination exhibited depression at only the higher filling pressures as compared with untreated or treated controls. The results suggest that functional alterations occurring as a result of diabetes cannot be accounted for by the depression of cardiac myosin ATPase and SR calcium uptake alone.

Diabetes-induced abnormalities in the myocardium.

One of the leading causes of mortality in diabetics is myocardial disease. In the past few years this subject has generated a significant amount of interest with the result that myocardial problems associated with diabetes are far better understood. Though originally thought to occur as a result of atherosclerosis, various studies have shown that heart disease can occur in the absence of atherosclerosis, suggesting a diabetic cardiomyopathy. Using diabetic animals, it has been possible to characterize diabetes-induced myocardial abnormalities. Diabetic rat hearts do not respond to conditions of high stress as well as controls. The functional depression is accompanied by altered cardiac enzyme systems. A decrease in myosin ATPase activity which appears to be a result of diabetes-induced hypothyroidism is seen. Also, a depression of sarcoplasmic reticular calcium ATPase, along with a depression of calcium uptake by the SR, is seen in diabetic rat hearts. Na+, K+ ATPase activity has also been shown to be depressed and the depression appears to correlate with depressed atrial contractility. High levels of circulating fats in diabetics may alter the integrity of membranes leading to altered enzyme activities. Insulin treatment has been relatively successful at reversing or preventing myocardial changes in the diabetic rat. Other treatments that have been studied include thyroid hormone treatment, since the depression of myosin ATPase can be corrected by such treatment; and carnitine treatment, as the elevation of long chain acyl carnitines (LCAC) and the resulting depression of calcium uptake in the SR can be so normalized. These treatments have not been successful at normalizing cardiac function. A combination of the two treatments normalized function only partially, suggesting that factors besides myosin ATPase and SR calcium uptake are involved. Other treatments that have been tried include vanadate, methyl palmoxirate, and choline and methionine. Vanadate treatment has proved to be encouraging in that it normalizes both function and hyperglycemia. Methyl palmoxirate, a fatty acid analog, normalized only the elevation of LCAC but did not affect function. Methionine and choline were only partially successful in preventing the functional alterations of diabetic rat hearts. The purpose of the present article is to review our understanding of diabetes-induced myocardial problems and their possible causes. Findings from our laboratory and others are described in which attempts have been made to normalize cardiac function.

Effects of insulin perfusion and altered glucose concentrations on heart function in 3-day and 6-week diabetic rats.

Diabetes is known to result in depression of myocardial function, whereas hearts from insulin-treated diabetic rats exhibit functional characteristics similar to controls. In the present study, we have studied the effect of insulin perfusion on cardiac performance of 3-day and 6-week streptozotocin (STZ) diabetic rats. Three days of diabetes did not result in depressed cardiac performance when the hearts were isolated and perfused in the working heart mode. Increasing the concentration of glucose from 5 to 10 mM in the perfusion fluid did not alter the function in either control or in diabetic rat hearts. However, when regular insulin or glucagon-free insulin (Humulin) (5 mU/mL) was included in the perfusion medium, the ventricular function of hearts from control rats was significantly enhanced, while diabetic myocardial function remained unaffected. When the study was repeated on hearts from 6-week diabetic animals, cardiac function of diabetic rats was significantly depressed as compared with controls. As in the 3-day study, contractility was not affected in either group by increasing glucose concentration in the perfusion medium. Again, inclusion of insulin in the medium enhanced cardiac contractility only in control hearts. These results suggest that diabetes results in a loss of myocardial sensitivity to insulin which seems to occur as early as 3 days after induction of diabetes with STZ. The study also demonstrates that the beneficial effects of in vivo insulin treatment on myocardial alterations induced by diabetes are not due to its direct myocardial effects.

A functional and ultrastructural analysis of experimental diabetic rat myocardium. Manifestation of a cardiomyopathy.

The effects of experimental diabetes on cardiac function and ultrastructure were studied in rats that had been diabetic for 6-24 wk. Experimental diabetes was produced by the intravenous (i.v.) injection of 65 mg/kg streptozocin (STZ) into rats 42-43 days old. Diabetic rat hearts perfused at 15 cm H2O on the working heart apparatus demonstrated depressed cardiac function (i.e., lower left ventricular pressure and +/- dP/dt) at 6, 12, and 24 wk of diabetes. Electron microscopic analysis of ventricular myocardium revealed increased lipid deposition from 6 to 24 wk of diabetes and progressive deterioration of the myocardial cell integrity at 12 and 24 wk of diabetes. This deterioration was characterized by loss of contractile protein, vacuolization (swollen sarcoplasmic reticulum), myelin formations, myocytolysis, and contracture bands. These alterations paralleled the depression of cardiac function at 12 and 24 wk of diabetes. There was, however, depressed function at 6 wk of diabetes but no observable alterations in myocardial ultrastructure. Therefore, experimental diabetes produced ultrastructural alterations in the rat heart that manifested themselves only after a demonstrable depression in cardiac function.

Prevention of diabetes-induced myocardial dysfunction in rats by methyl palmoxirate and triiodothyronine treatment.

Diabetes results in myocardial functional alterations which are accompanied by a depression of biochemical parameters such as myosin ATPase and calcium uptake in the sarcoplasmic reticulum. Methyl palmoxirate, a fatty acid analog, is reported to decrease circulating glucose levels by inhibiting fatty acid metabolism, thus forcing carbohydrate utilization. In the present study, we attempted to prevent streptozotocin diabetes-induced myocardial alterations in the rat. Using the isolated working heart preparation, we observed a depression of myocardial function in rats 6 weeks after the induction of diabetes, which was characterized by the inability of these hearts to develop left ventricular pressures and rates of ventricular contraction and relaxation as well as control hearts at higher left atrial filling pressures. Methyl palmoxirate treatment (25 mg kg-1 day-1 po daily) was unable to control diabetes-induced changes in plasma glucose, triglycerides, insulin, and total lipids. Also, the functional depression seen in diabetic rat hearts was present despite the treatment. However, depression of calcium uptake and elevation of long chain acyl carnitines seen in sarcoplasmic reticulum (SR) prepared from diabetic rat hearts could be prevented by the treatment. As triiodothyronine (T3) treatment has been shown to normalize depression of cardiac myosin ATPase in diabetic rats, we repeated the study using a combination of T3 (30 micrograms kg-1 day-1 sc daily) and methyl palmoxirate. While diabetic rats treated with T3 alone did not show significant improvement of myocardial function when compared with untreated diabetics, the function of those treated with both T3 and methyl palmoxirate was not significantly different from that in control rat hearts.(ABSTRACT TRUNCATED AT 250 WORDS)

Effect of vanadate on elevated blood glucose and depressed cardiac performance of diabetic rats.

The trace element vanadium has an unclear biological function. Vanadate, an oxidized form of vanadium, appears to have an insulin-like action. The effect of vanadate on blood glucose and cardiac performance was assessed in female Wistar rats 6 weeks after they were made diabetic with streptozotocin. When vanadate was administered for a 4-week period to the diabetic rats, their blood glucose was not significantly different from that of nondiabetic controls despite a low serum insulin. In contrast, blood glucose was increased about threefold in the diabetic rats that were not treated with vanadate; these rats also had low insulin levels. Cardiac performance was depressed in the untreated diabetic animals, but the cardiac performance of the vanadate-treated diabetic animals was not significantly different from that of nondiabetic controls. Thus vanadate controlled the high blood glucose and prevented the decline in cardiac performance due to diabetes.

Lack of effect of thyroid hormone on diabetic rat heart function and biochemistry.

Cardiac functional abnormalities are frequently seen in diabetics and diabetes is also known to produce a state of mild hypothyroidism. To study the degree of involvement of diabetes-induced hypothyroidism on altered myocardial function, thyroid replacement therapy was carried out in streptozotocin-diabetic rats. Triiodothyronine (T3) treatment was initiated 3 days after the rats were made diabetic and was carried out for 6 weeks thereafter. Isolated perfused hearts from diabetic rats exhibited a depression in left ventricular developed pressure and positive and negative dP/dt at higher filling pressures as compared with controls. The depression could not be prevented by thyroid treatment. Calcium uptake activity in the cardiac sarcoplasmic reticulum (SR) was also depressed as a result of diabetes and this depression also was not prevented by thyroid treatment. Long chain acyl carnitine levels were found to be elevated in diabetic cardiac SR and could not be lowered by T3 treatment. The results indicate that the myocardial dysfunction observed in diabetic rats is due to factors other than the induced hypothyroidism.

Effect of insulin treatment on long-term diabetes-induced alteration of myocardial function.

Six months following the induction of diabetes by streptozotocin (50 mg/kg i.v.) diabetic rats exhibited elevated levels of plasma glucose and glycosylated hemoglobin. Plasma insulin levels were 50% of control and diabetic animals weighed significantly less than control. Using a working heart preparation it was found that (+) and (-) dP/dt and left ventricular pressure development (LVDP) was decreased in hearts from diabetic animals. Insulin treatment (9 U/kg/day s.c. of protamine zinc insulin) for 4 weeks prior to sacrifice restored body weight and plasma insulin to normal. Plasma glucose and glycosylated hemoglobin levels were significantly decreased towards normal in insulin treated diabetic rats. LVDP and (+) dP/dt was also partially returned to normal in insulin treated diabetic rats while (-) dP/dt was completely reversed to normal. Thus, 4 weeks of insulin treatment to rats previously diabetic for 5 months partially or totally reversed the changes produced by diabetes.

Cardiac sarcoplasmic reticulum function in insulin- or carnitine-treated diabetic rats.

Cardiac sarcoplasmic reticulum (SR) function and SR levels of long-chain (LC) acylcarnitines were determined in streptozotocin-induced diabetic rats treated with insulin or D,L-carnitine. ATP-dependent calcium transport was significantly depressed in cardiac SR isolated from untreated diabetic rats compared with control rats. Diabetic rat cardiac SR levels of LC acylcarnitines were also significantly elevated. Various parameters of heart function (left ventricular developed pressure, +dP/dT, and -dP/dT), as determined on an isolated working heart apparatus, were found to be depressed in untreated diabetic rats. Cardiac SR isolated from diabetic rats treated throughout the study period with insulin or D,L-carnitine did not have elevated levels of LC acylcarnitines associated with SR membrane nor was SR calcium transport activity depressed. Heart function in the diabetic rats treated with insulin was similar to control rat hearts but heart function remained depressed in diabetic rats treated with D,L-carnitine. The data suggest that the LC acylcarnitines are involved in the observed impairment of cardiac SR function in diabetic rats. Other factors, however, must be contributing to the depression in heart function noted in these animals.

Prevention and reversal of altered myocardial function in diabetic rats by insulin treatment.

Isolated perfused hearts from diabetic rats exhibit a decreased responsiveness to increasing work loads. However, the precise time point at which functional alterations occur is not clearly established. Previous observations in our laboratory have suggested that the alterations in myocardial function are not apparent at 30 days whereas they are clearly seen 100 days after streptozotocin-induced diabetes. We studied the cardiac function of 6-week diabetic rats using the isolated perfused heart preparation. The 6-week time period was found to be sufficient to cause depression of myocardial function in these animals. We also studied the effect of insulin treatment on myocardial performance of diabetic rats. Insulin treatment was initiated 3 days and 6 weeks after injection of streptozotocin (STZ). The treatment was continued for 6 and 4 weeks in the respective groups. Hearts from 6-week diabetic animals exhibited a depressed left ventricular developed pressure (LVDP) and positive and negative dP/dt at higher filling pressures when compared with 6-week control animals. However, the depression was not seen in the 6-week insulin-treated diabetic animals. Ten-week diabetic rat hearts also showed a depression of LVDP and positive and negative dP/dt when compared with 10-week controls. The group of animals that had been diabetic for 6 weeks and then treated for 4 weeks with insulin exhibited a reversal of the depressed myocardial function. These results demonstrate that depression of myocardial performance, which is evident 6 weeks after diabetes is induced, can be prevented if insulin treatment is initiated as the disease is induced. Further, insulin treatment is capable of reversing the abnormalities after they have occurred.

Continuous long-term insulin delivery in diabetic rats utilizing implanted osmotic minipumps.

Surgically implanted osmotic minipumps were used to continuously deliver insulin to chemically-induced diabetic rats. Serum glucose levels were maintained within normal limits for 7 days in all diabetic rats implanted with the minipumps. Beyond this time period, serum glucose levels could not be adequately controlled in greater than 50% of the diabetic rats. Seven of 20 diabetic rats originally implanted with osmotic minipumps were well controlled throughout the 31-day study period. This study demonstrates that insulin-filled osmotic minipumps provide no advantage over daily insulin injection in the long term control of diabetes.