513
Nonesterified Fatty Acids in Normal and Diabetic Rat Sciatic Nerve Jyotiprakas Chattopadhyay, Ed W. Thompson and Harald H.O. Schmid* The Hormel Institute, University of Minnesota, Austin, Minnesota 55912
Alloxan-induced diabetes in rats results in elevated levels of nonesterified fatty acids (NEFA) in whole sciatic nerve and its endoneurium. Increases in N E F A levels are more pronounced in whole diabetic nerve (40% over control) than in its endoneurial portion (20-30%). Alterations in the composition of phospholipid fatty acids are observed as well, including an increase in linoleate (18:2n-6) in endoneurial phosphatidylethanolamine and a decrease in arachidonate (20:4n4}) in both phosphatidylethanolamine and phosphatidylinositol of diabetic nerve. Lipids 27, 513-517 (1992).
Peripheral neuropathy is a common complication of human diabetes meUitus as well as of chemically-induced or genetically determined diabetes in experimental animals (1-3). Common functional changes induced by diabetes in peripheral nerve include reduced conduction velocity, action potential and axonal transport, whereas morphological changes can include segmental demyelination (2). The underlying physiological and biochemical alterations induced by diabetes appear to be related to the microenvironment of the nerve, including blood flow and oxygen delivery (4) and changes in myo-inositol and inositol phospholipid metabolism (5). Reduced blood flow and hypoxia observed in diabetic rat sciatic nerve (6) were recently correlated with the generation of oxygen free radicals and the accumulation of lipid peroxidation products (7). Ischemia and hypoxia are often associated with the liberation of fatty acids from complex lipids by the action of lipolytic enzymes, a process which is especially pronounced in the central nervous system (8). In addition, alterations in fatty acid metabolism are an important component of diabetes mellitus (9) and we have previously observed significantly elevated levels of nonesterified fatty acids (NEFA) in the myocardium of alloxan-diabetic rats (10). There exists, however, very little information on the amount or composition of NEFA in peripheral nerve in general (11-13) and none regarding diabetic animals. In order to examine whether changes in nonesterified fatty acids may play a role in the development of diabetic neuropathy, we have measured their levels and composition in the endoneurium and in whole sciatic nerve from normal and aUoxan diabetic rats. The results are reported hem
EXPERIMENTAL PROCEDURES Experimental animals and induction of diabetes meUitus. Male Sprague-Dawley rats of 150-180 g were obtained from a commercial colony (Bio Lab Corporation, St. Paul, MN) and housed in separate cages within a modern animal
*To whom correspondence should be addressed. Abbreviations: GC, gas chromatography; NEFA, nonesterifiedfatty acids; TLC, thin-layer chromatography.
care facility. All animals had free access to food (Purina #5001, St. Louis, MO) and water throughout the study and were exposed to a 12 h light/12 h dark cycle Diabetes was induced in each animal by a single injection of 4% alloxan monohydrate (Sigma Chemical Company, St. Louis, MO) in sterile saline into a taft vein at a dosage of 55 mg/kg body weight as described previously (10,14). Control animals received the appropriate volume of sterile saline Nonfasting plasma glucose levels were measured three days later and at monthly intervals by enzymatic assay (Sigma) coupling the reactions of glucose oxidase and peroxidase Glycosylated hemoglobin was measured according to the method of Abraham et al. (15) using Isolab affinity columns (Isolal~ Inn, Akron, OH). Alloxaninjected rats with plasma glucose levels greater than 300 mg/dL and showing little or no weight gain were considered insulin deficient, and saline-injected animals with glucose levels below 150 mg/dL and normal rates of weight gain were used as controls. Any animal not meeting those criteria (i.~, with plasma glucose level between 150 and 300 mg/dL or hyperglycemic but weight gaining) was excluded from the study. These limited criteria for inclusion of animals, used routinely in our studies of experimental diabetes, were selected (10,14) to minimize experimental variations due to differing severities of diabetes which we have shown to affect myocardial complications (16). In accord with standards of proper animal care" efforts were made throughout the study to mi~imlze or eliminate any unnecessary discomfort to the experimental animals. A diabetic rat of this model will normally remain active and able to groom itself. Food and water consumption and urine output will be above normal. In these experiments, any animal which showed a sudden or significant chronic weight loss, developed anuria, or became lethargic or unable to eat or drink was removed from the study and sacrificed by COs inhalation. At the intervals described below, diabetic and nondiabetic rats were moderately anesthetized by subcutaneous injection of sodium pentobarbital (50 mg/kg body weight) and brought to full anesthesia by intraperitoneal injection at the same dosage. The skin and superficial fascia of the right thigh of each rat were incised to expose the sciatic nerve from the point where it emerged from the gluteal muscles, past its separation into tibial, peroneal, and sural divisions, to the points where these passed deep to peripheral musculature This portion was quickly excised and prepared for lipid analysis as described below. The animal was then killed by decapitation while still under anesthesia Preparation of sciatic nerve tissue. Experiment 1: Agematched diabetic and nondiabetic control animals were killed 8 wk after the induction of diabetes. The excised sciatic nerve was washed in saline on cooled (15~ dental wax, freed of adherent blood vessels and connective tissue, frozen in liquid N2, weighed, and prepared for lipid analysis. Experiment 2: Age-matched diabetic and nondiabetic control animals were killed 21 wk after induction of diabetes. Upon removal, the nerve was washed in cooled saline, and the epineurium and perineurium were LIPIDS, Vol. 27, no. 7 (1992)
514
J. CHATTOPADHYAYET AL. stripped away by the procedure of Dyck et aL (17). The remaining endoneurium was weighed and prepared for lipid analysis. Analysis of sciatic nerve lipids. Total lipids were extracted from sciatic nerve endoneurium or total sciatic nerve according to the procedure of Folch et at (18). Briefly, nerve tissue was homogenized in cold methanol containing butylated hydroxytoluene using a Dounce homogenizer and then extracted with chloroform/methanol (2:1, vol/vol) containing 10 ~g of beptadecanoic acid as internal standard. The phases were separated by adcling 0.2 vol of 0.9% NaCI to the mixture which was again vortexed to remove water soluble contaminants. The residual tissue and upper aqueous layer were re-extracted with chloroform]methanol/water (86:14:1, vol]vol/vol). The pooled chloroform layer was evaporated under N 2 and the lipids redissolved in chloroform for further analysis. Aliquots were assayed for lipid phosphorus according to B a r t l e t t (19). A portion of the lipid extract was analyzed for N E F A after addition of diazomethane (20) using beptadecanoic acid as an internal standard. Briefly, the lipid extract was dried down under N2, dissolved in freshly prepared diethyl ether/methanol (9:1, vol/vol), and treated with a solution of diazomethane in diethyl ether for 10-15 mino F a t t y acid methyl esters were subsequently purified by silicic acid column chromatography and analyzed by gas chromatography using a Packard (Downers Grove, IL) model 428 gas chromatograph equipped with a Supelco (Bellefonte, PA) 2330 packed column and flame ionization detector. Column temperature was programmed from 180~ to 235~ at 2~ and the detector and injection port temperatures were 260~ and 250~ respectively. The peaks were identified by comparison with standards obtained from NuChek Prep (Elysian, MN) and aliquots were hydrogenated to confirm the identity of unsaturated f a t t y acids. Peak areas were quantitated with a Spectra-Physice (San Jose, CA) SP4270 computing integrator. Another portion of the lipid extract was applied to a small column of silicic acid and eluted with 10 ml, chloroform (neutral lipids), 20 mL acetone (glycolipids) and 10 m L methanol (phospholipids). The phospholipid classes were quantified by phosphorus assay (19) after separation by two-dimensional thin-layer chromatography (TLC) {21) as described previously (22). For the analysis of their constituent f a t t y acids, phospholipid classes were isolated by one-dimensional TLC on layers of silica gel H containing 7.5% magnesium acetate using chloroform/methanol/ammonium hydroxide/water (65:35:5:1, by vol) as developing solvent. Choline and ethanolamine phospholipids were ftm ther purified by TLC on silica gel H using chloroform/ methanol/water (65:25:4, vol/vol/vol). Ethanolamine plasmalogens were hydrolyzed in HCl-saturated diethyl ether, and the resulting lysophosphatidylethanolamines were separated from the nonhydrolyzed diacyl analogs by TLC. Purified phospholipids were transesterified with 0.2 N N a O H in methanol at 45~ for 1 h. Methyl esters were extracted into hexane and analyzed by gas chromatography (GC) as described above Statistics. All values between diabetic and non-diabetic groups in Experiments 1 and 2 were compared by twotailed, unpaired Student's t-test, with significance indicated by P < 0.05. LIPIDS, Vol. 27, no. 7 (1992)
RESULTS
Effects of alloxan-induced diabetes. As demonstrated b y the data of Table 1, diabetic animals exhibited severe hyperglycemia, elevated levels of glycesylated hemoglobin and plasma lipids, increased water intake and essentially no weight gain compared to nondiabetic control rats. The wet weight of the endoneurium was reduced in diabetic animals, b u t increased moderately when corrected for body weight due to the marked loss of adipose tissue (14,16} which offsets continued growth of lean body mass to yield little or no net weight gain in this model. NE F A content and composition of sciatic nerve. As shown in Table 2, Experiment 1 demonstrated t h a t the total amount of sciatic nerve N E F A was significantly higher in diabetic rats t h a n in age-matched non-diabetic control animals when expressed relative to either wet TABLE 1 Effects of AIIoxan-lnduced Diabetes a
Body weight (g) Weight gain (g/wk) Sciatic endoneurium, wet wt (rag) Sciatic endoneurium, wet wt/body wt (mg/g) Water intake (mL/day) Plasma lipid P (~mo]/mL) Glycosylated hemoglobin (%) Blood glucose (mg/dL)
Nondiabetic control
Diabetic
625 + 29 20.6 + 1.1
240 + 63b 1.7 + 3.1 b
30.2 + 4.6
19.8 + 4.7 b
0.049 + 0.008 49.8 + 11.1 0.97 + 0.10
0.085 + 0.019b 280.9 + 68.8b 1.67 + 0.34b
4.98 + 0.64 108.3 + 7.1
16.01 + 2.60b 533.3 + 65.0b
aMeasured 21 wk after administration of alloxan (55 mg/kg body wt, i.v.) or saline to 6 wk old rats. Each value represents ~ + SD of 4 control and 7 diabetic Animals. bSigniflcant differencebetween control and diabetic group (P ~ 0.05). TABLE 2 Content and Composition of Nonesterified Fatty Acids (NEFA) of Whole Sciatic Nerve from Nondiabetic and Diabetic Rats a
nmol NEFA per ~anol lipid P /~mol NEFA per g wet tissue
Nondlabetic normal
Diabetic
14.8 + 1.9
21.1 + 1.6b
1.22 + 0.13
2.23 + 0.31b mol % FA
16:0 16:1 18:0 18:1 18:2n-6 20:0 20:4n-6 22:0 24:0
33.3 + 3.0 + 14.2 + 37.6 + 2.1 + 1.6 + 2.8 + 3.1 + 2.4 +
0.8 0.4 1.3 0.3 0.1 0.1 0.4 0.3 0.3
32.4 + 1.5 2.1 + 0.02b 15.0 + 1.2 38.5 + 2.1 2.8 +- 0.5b 1.6 + 0.1 2.4 _ 0.2 3.1 --- 0.2 1.9 + 0.2
aAge, 14.5 wk; duration of diabetes, 8 wk. Each value represents + SD of 4 nerve samples. bsim~iflcant differencebetween control and diabetic group {P ~ 0.05).
515
NONESTERIFIED FATTY ACIDS IN NERVE weight or total lipid phosphorus. When the N E F A composition of the nerve was examined, a modest decrease in pelmitoleic acid (16:1) and increase in linoleic acid (18:2n-6) were noted in nerve from diabetic animals. The levels of arachidonic acid (20:4n-6) and tetracosanoic acid (24:0) were also reduced although these changes were not statisticatly significant.
k"- O3 O0 r
k
Nonesterified Fatty Acids in Normal and Diabetic Rat Sciatic Nerve Jyotiprakas Chattopadhyay, Ed W. Thompson and Harald H.O. Schmid* The Hormel Institute, University of Minnesota, Austin, Minnesota 55912
Alloxan-induced diabetes in rats results in elevated levels of nonesterified fatty acids (NEFA) in whole sciatic nerve and its endoneurium. Increases in N E F A levels are more pronounced in whole diabetic nerve (40% over control) than in its endoneurial portion (20-30%). Alterations in the composition of phospholipid fatty acids are observed as well, including an increase in linoleate (18:2n-6) in endoneurial phosphatidylethanolamine and a decrease in arachidonate (20:4n4}) in both phosphatidylethanolamine and phosphatidylinositol of diabetic nerve. Lipids 27, 513-517 (1992).
Peripheral neuropathy is a common complication of human diabetes meUitus as well as of chemically-induced or genetically determined diabetes in experimental animals (1-3). Common functional changes induced by diabetes in peripheral nerve include reduced conduction velocity, action potential and axonal transport, whereas morphological changes can include segmental demyelination (2). The underlying physiological and biochemical alterations induced by diabetes appear to be related to the microenvironment of the nerve, including blood flow and oxygen delivery (4) and changes in myo-inositol and inositol phospholipid metabolism (5). Reduced blood flow and hypoxia observed in diabetic rat sciatic nerve (6) were recently correlated with the generation of oxygen free radicals and the accumulation of lipid peroxidation products (7). Ischemia and hypoxia are often associated with the liberation of fatty acids from complex lipids by the action of lipolytic enzymes, a process which is especially pronounced in the central nervous system (8). In addition, alterations in fatty acid metabolism are an important component of diabetes mellitus (9) and we have previously observed significantly elevated levels of nonesterified fatty acids (NEFA) in the myocardium of alloxan-diabetic rats (10). There exists, however, very little information on the amount or composition of NEFA in peripheral nerve in general (11-13) and none regarding diabetic animals. In order to examine whether changes in nonesterified fatty acids may play a role in the development of diabetic neuropathy, we have measured their levels and composition in the endoneurium and in whole sciatic nerve from normal and aUoxan diabetic rats. The results are reported hem
EXPERIMENTAL PROCEDURES Experimental animals and induction of diabetes meUitus. Male Sprague-Dawley rats of 150-180 g were obtained from a commercial colony (Bio Lab Corporation, St. Paul, MN) and housed in separate cages within a modern animal
*To whom correspondence should be addressed. Abbreviations: GC, gas chromatography; NEFA, nonesterifiedfatty acids; TLC, thin-layer chromatography.
care facility. All animals had free access to food (Purina #5001, St. Louis, MO) and water throughout the study and were exposed to a 12 h light/12 h dark cycle Diabetes was induced in each animal by a single injection of 4% alloxan monohydrate (Sigma Chemical Company, St. Louis, MO) in sterile saline into a taft vein at a dosage of 55 mg/kg body weight as described previously (10,14). Control animals received the appropriate volume of sterile saline Nonfasting plasma glucose levels were measured three days later and at monthly intervals by enzymatic assay (Sigma) coupling the reactions of glucose oxidase and peroxidase Glycosylated hemoglobin was measured according to the method of Abraham et al. (15) using Isolab affinity columns (Isolal~ Inn, Akron, OH). Alloxaninjected rats with plasma glucose levels greater than 300 mg/dL and showing little or no weight gain were considered insulin deficient, and saline-injected animals with glucose levels below 150 mg/dL and normal rates of weight gain were used as controls. Any animal not meeting those criteria (i.~, with plasma glucose level between 150 and 300 mg/dL or hyperglycemic but weight gaining) was excluded from the study. These limited criteria for inclusion of animals, used routinely in our studies of experimental diabetes, were selected (10,14) to minimize experimental variations due to differing severities of diabetes which we have shown to affect myocardial complications (16). In accord with standards of proper animal care" efforts were made throughout the study to mi~imlze or eliminate any unnecessary discomfort to the experimental animals. A diabetic rat of this model will normally remain active and able to groom itself. Food and water consumption and urine output will be above normal. In these experiments, any animal which showed a sudden or significant chronic weight loss, developed anuria, or became lethargic or unable to eat or drink was removed from the study and sacrificed by COs inhalation. At the intervals described below, diabetic and nondiabetic rats were moderately anesthetized by subcutaneous injection of sodium pentobarbital (50 mg/kg body weight) and brought to full anesthesia by intraperitoneal injection at the same dosage. The skin and superficial fascia of the right thigh of each rat were incised to expose the sciatic nerve from the point where it emerged from the gluteal muscles, past its separation into tibial, peroneal, and sural divisions, to the points where these passed deep to peripheral musculature This portion was quickly excised and prepared for lipid analysis as described below. The animal was then killed by decapitation while still under anesthesia Preparation of sciatic nerve tissue. Experiment 1: Agematched diabetic and nondiabetic control animals were killed 8 wk after the induction of diabetes. The excised sciatic nerve was washed in saline on cooled (15~ dental wax, freed of adherent blood vessels and connective tissue, frozen in liquid N2, weighed, and prepared for lipid analysis. Experiment 2: Age-matched diabetic and nondiabetic control animals were killed 21 wk after induction of diabetes. Upon removal, the nerve was washed in cooled saline, and the epineurium and perineurium were LIPIDS, Vol. 27, no. 7 (1992)
514
J. CHATTOPADHYAYET AL. stripped away by the procedure of Dyck et aL (17). The remaining endoneurium was weighed and prepared for lipid analysis. Analysis of sciatic nerve lipids. Total lipids were extracted from sciatic nerve endoneurium or total sciatic nerve according to the procedure of Folch et at (18). Briefly, nerve tissue was homogenized in cold methanol containing butylated hydroxytoluene using a Dounce homogenizer and then extracted with chloroform/methanol (2:1, vol/vol) containing 10 ~g of beptadecanoic acid as internal standard. The phases were separated by adcling 0.2 vol of 0.9% NaCI to the mixture which was again vortexed to remove water soluble contaminants. The residual tissue and upper aqueous layer were re-extracted with chloroform]methanol/water (86:14:1, vol]vol/vol). The pooled chloroform layer was evaporated under N 2 and the lipids redissolved in chloroform for further analysis. Aliquots were assayed for lipid phosphorus according to B a r t l e t t (19). A portion of the lipid extract was analyzed for N E F A after addition of diazomethane (20) using beptadecanoic acid as an internal standard. Briefly, the lipid extract was dried down under N2, dissolved in freshly prepared diethyl ether/methanol (9:1, vol/vol), and treated with a solution of diazomethane in diethyl ether for 10-15 mino F a t t y acid methyl esters were subsequently purified by silicic acid column chromatography and analyzed by gas chromatography using a Packard (Downers Grove, IL) model 428 gas chromatograph equipped with a Supelco (Bellefonte, PA) 2330 packed column and flame ionization detector. Column temperature was programmed from 180~ to 235~ at 2~ and the detector and injection port temperatures were 260~ and 250~ respectively. The peaks were identified by comparison with standards obtained from NuChek Prep (Elysian, MN) and aliquots were hydrogenated to confirm the identity of unsaturated f a t t y acids. Peak areas were quantitated with a Spectra-Physice (San Jose, CA) SP4270 computing integrator. Another portion of the lipid extract was applied to a small column of silicic acid and eluted with 10 ml, chloroform (neutral lipids), 20 mL acetone (glycolipids) and 10 m L methanol (phospholipids). The phospholipid classes were quantified by phosphorus assay (19) after separation by two-dimensional thin-layer chromatography (TLC) {21) as described previously (22). For the analysis of their constituent f a t t y acids, phospholipid classes were isolated by one-dimensional TLC on layers of silica gel H containing 7.5% magnesium acetate using chloroform/methanol/ammonium hydroxide/water (65:35:5:1, by vol) as developing solvent. Choline and ethanolamine phospholipids were ftm ther purified by TLC on silica gel H using chloroform/ methanol/water (65:25:4, vol/vol/vol). Ethanolamine plasmalogens were hydrolyzed in HCl-saturated diethyl ether, and the resulting lysophosphatidylethanolamines were separated from the nonhydrolyzed diacyl analogs by TLC. Purified phospholipids were transesterified with 0.2 N N a O H in methanol at 45~ for 1 h. Methyl esters were extracted into hexane and analyzed by gas chromatography (GC) as described above Statistics. All values between diabetic and non-diabetic groups in Experiments 1 and 2 were compared by twotailed, unpaired Student's t-test, with significance indicated by P < 0.05. LIPIDS, Vol. 27, no. 7 (1992)
RESULTS
Effects of alloxan-induced diabetes. As demonstrated b y the data of Table 1, diabetic animals exhibited severe hyperglycemia, elevated levels of glycesylated hemoglobin and plasma lipids, increased water intake and essentially no weight gain compared to nondiabetic control rats. The wet weight of the endoneurium was reduced in diabetic animals, b u t increased moderately when corrected for body weight due to the marked loss of adipose tissue (14,16} which offsets continued growth of lean body mass to yield little or no net weight gain in this model. NE F A content and composition of sciatic nerve. As shown in Table 2, Experiment 1 demonstrated t h a t the total amount of sciatic nerve N E F A was significantly higher in diabetic rats t h a n in age-matched non-diabetic control animals when expressed relative to either wet TABLE 1 Effects of AIIoxan-lnduced Diabetes a
Body weight (g) Weight gain (g/wk) Sciatic endoneurium, wet wt (rag) Sciatic endoneurium, wet wt/body wt (mg/g) Water intake (mL/day) Plasma lipid P (~mo]/mL) Glycosylated hemoglobin (%) Blood glucose (mg/dL)
Nondiabetic control
Diabetic
625 + 29 20.6 + 1.1
240 + 63b 1.7 + 3.1 b
30.2 + 4.6
19.8 + 4.7 b
0.049 + 0.008 49.8 + 11.1 0.97 + 0.10
0.085 + 0.019b 280.9 + 68.8b 1.67 + 0.34b
4.98 + 0.64 108.3 + 7.1
16.01 + 2.60b 533.3 + 65.0b
aMeasured 21 wk after administration of alloxan (55 mg/kg body wt, i.v.) or saline to 6 wk old rats. Each value represents ~ + SD of 4 control and 7 diabetic Animals. bSigniflcant differencebetween control and diabetic group (P ~ 0.05). TABLE 2 Content and Composition of Nonesterified Fatty Acids (NEFA) of Whole Sciatic Nerve from Nondiabetic and Diabetic Rats a
nmol NEFA per ~anol lipid P /~mol NEFA per g wet tissue
Nondlabetic normal
Diabetic
14.8 + 1.9
21.1 + 1.6b
1.22 + 0.13
2.23 + 0.31b mol % FA
16:0 16:1 18:0 18:1 18:2n-6 20:0 20:4n-6 22:0 24:0
33.3 + 3.0 + 14.2 + 37.6 + 2.1 + 1.6 + 2.8 + 3.1 + 2.4 +
0.8 0.4 1.3 0.3 0.1 0.1 0.4 0.3 0.3
32.4 + 1.5 2.1 + 0.02b 15.0 + 1.2 38.5 + 2.1 2.8 +- 0.5b 1.6 + 0.1 2.4 _ 0.2 3.1 --- 0.2 1.9 + 0.2
aAge, 14.5 wk; duration of diabetes, 8 wk. Each value represents + SD of 4 nerve samples. bsim~iflcant differencebetween control and diabetic group {P ~ 0.05).
515
NONESTERIFIED FATTY ACIDS IN NERVE weight or total lipid phosphorus. When the N E F A composition of the nerve was examined, a modest decrease in pelmitoleic acid (16:1) and increase in linoleic acid (18:2n-6) were noted in nerve from diabetic animals. The levels of arachidonic acid (20:4n-6) and tetracosanoic acid (24:0) were also reduced although these changes were not statisticatly significant.
k"- O3 O0 r
k
