[Frontiers in Bioscience 3, d169-175, February 15, 1998]

	[Frontiers in Bioscience 3, d169-175, February 15, 1998] Reprints PubMed CAVEAT LECTOR
	FREE FATTY ACIDS (FFA), A LINK BETWEEN OBESITY AND INSULIN RESISTANCE Guenther Boden, M.D. Division of Endocrinology/Diabetes/Metabolism and the General Clinical Research Center, Temple University School of Medicine, Philadelphia, PA Received 1/21/98 Accepted 1/31/98 4. FFA AND PERIPHERAL INSULIN RESISTANCE 4.1. Effect of FFA on Basal Glucose Uptake When plasma FFA concentrations were increased acutely by IV infusion of a triglyceride emulsion (Liposyn II, 10% safflower and 10% soybean oil) and heparin (which enhances lipoprotein lipase activity), FFA levels rose 2-3 fold (to between 1.0 and 1.5 mM) but basal rates of glucose uptake did not change (12). 4.2 Effect of FFA on Insulin Stimulated Glucose Uptake In healthy subjects, acute elevation of plasma FFA by IV lipid/heparin infusion inhibited total body glucose uptake dose dependently after a lag period of 3-4 hours "figure 1". Inhibition of CHO oxidation occurred ~ 2 h earlier than the inhibition of glucose uptake. The latter was reversed ~ 3 hours after discontinuation of lipid/heparin infusions (13) "figure 2". Other studies showed that the lipid induced inhibition of insulin stimulated glucose uptake was linear (14,15) "figure 3" and occurred similarly in healthy subjects and in patients with Type II diabetes (14-20). The long delay of 3-4 hours between the start of the insulin and lipid infusions and the development of significant insulin resistance was the most likely reason why the inhibitory effect of FFA on glucose uptake was not found in many previous studies (21-25). Figure 1: Effect of plasma FFA on insulin-stimulated glucose uptake. Euglycemic-hyperinsulinemic (~ 420 pmol/l) clamping was performed in healthy volunteers for 6 h. High levels of plasma FFA were produced by the infusion of triglycerides (4.3 mmol/min) plus heparin (0.4 U A kg^-1 . min^-1) (D , n = 4); intermediate plasma FFA levels, by infusion of triglycerides without heparin (o, n = 4); and low FFA levels, by infusion of saline alone (·, n = 6). Data shown are mean ± SE. The inhibition of insulin-stimulated glucose uptake became statistically significant ~ 3.5 h after the start of the lipid infusion. p < 0.05; p < 0.01, comparing high with low FFA. Adapted from Boden et al. (14). Figure 2. Plasma FFA levels, and rates of CHO oxidation (COX) and glucose disappearance (G_Rd) in healthy men during lipid/heparin infusion (solid circles) and during lipid/heparin infusion from 0-120 min followed by saline infusion from 120-360 min (open circles) * p < 0.05, p < 0.01, * p < 0.005 comparing the 2 groups. Adapted from Boden et al. (13). The linear relationship between plasma FFA levels and insulin stimulated glucose uptake in healthy controls and patients with Type II diabetes suggested that for every 100 mM increase in plasma FFA, peripheral insulin sensitivity decreased by ~ 8% "figure 3". It needs to be emphasized, however, that the total range of insulin stimulated glucose uptake was approximately 2 fold higher in healthy controls than in patients with Type II diabetes (15) "figure 4". Thus, FFA could account for maximally 50% of peripheral insulin resistance in patients with Type II diabetes. Several potential problems could affect the interpretation of these results. First, the infused triglyceride emulsion (Liposyn II) contained a considerable amount of glycerol (2.5 grams/100 ml). We have shown, however, that this amount of glycerol did not affect glucose uptake (15). Second, the data showing that elevation of plasma FFA decreased peripheral glucose uptake in response to insulin were all produced in acute experiments. Their relevance relative to the long term effects of elevated plasma FFA, for instance during obesity could, therefore, be questioned. It was, however, shown that lowering of plasma FFA below basal values caused an increase in insulin stimulated glucose uptake (13) "figure 1". This indicated that basal FFA levels exerted long term inhibitory effects on peripheral glucose uptake. Figure 3.* Relationship between insulin stimulated glucose uptake and plasma FFA in 7 patients with NIDDM and in 6 non-diabetic controls. Data from controls were obtained during euglycemic (~ 5 mM)-hyperinsulinemic (~ 450 pM) clamping. Data from patients with NIDDM were obtained during hyperinsulinemic (~ 11 mM)-hyperinsulinemic clamping. Adapted from Boden et al. (14,15). Figure 4.* Insulin-stimulated glucose uptake at comparable low plasma FFA (< 100 mmol/l) and euglycemia in diabetic and nondiabetic subjects. Shown are insulin-stimulated glucose uptakes before (g) and after ( ) 4 h of euglycemic (~ 4.8 mmol/l) hyperinsulinemic (~ 500 pmol/l) clamping in 7 patients with NIDDM and 6 nondiabetic control subjects. Preclamp glucose uptake could not be obtained in the diabetic patients because insulin was infused to lower their blood glucose concentrations into the normal range. Insulin-stimulated glucose uptake (G_Rd) was ~ 2 times higher in nondiabetic than in diabetic individuals (30 vs. 58 mmol/kg fat free mass, p < 0.01). Triglyceride plus heparin infusion (P) decreased insulin-stimulated glucose uptake by ~50% in diabetic and nondiabetic individuals. EU, euglycemia; FFM, fat free mass. Adapted from Boden et al. (14,15). 4.3 Cellular Location of FFA Induced Defects* To determine the cellular location of the fat induced inhibition of insulin stimulated glucose uptake, glucose fluxes through all major pathways of intracellular glucose utilization were determined non-invasively. Rates of glucose uptake and glycolysis were estimated with 3-³H-glucose and glycogen synthesis was obtained by subtracting glycolysis from glucose uptake. The validity of these methods has been validated (26). CHO oxidation was determined by indirect calorimetry and non-oxidative glycolysis (lactate and alanine formation) by subtracting CHO oxidation from glycolysis. Using these methods, it was found that lipid/heparin infusion inhibited rates of glucose uptake, glycogen synthesis and glycolysis about equally (15). In normal controls, for instance, plasma FFA concentrations of ~ 600 mM inhibited insulin stimulated glucose uptake, glycogen synthesis and glycolysis all by ~ 50% while in patients with Type II diabetes, a higher plasma FFA concentration of ~ 1200 mM resulted in a ~ 90% inhibition "figure 5". These results suggested a FFA induced defect at the level of transport and/or phosphorylation (the methods used could not differentiate between these two possibilities). A primary defect at the level of glycogen synthesis or glycolysis, on the other hand, would have produced disproportionate reductions in the flux rates through these pathways. Figure 5. Effects of elevated plasma FFA on rates of glucose uptake (G_Rd), glycogen synthesis (GS), and glycolysis (GLS) in 7 patients with NIDDM during hyperinsulinemic (~ 900 pmol/l) isoglycemic (~ 11 mmol/l) clamping and in 6 nondiabetic control subjects during euglycemic-hyperinsulinemic (~ 500 pmol/l) clamping. Total length of bars represent insulin-stimulated G_Rd , GS, or GLS, set as 100%. The darkly shaded parts of the bars represent insulin-stimulated G_Rd , GS, or GLS after 4 h of elevated plasma FFA (~ 1,200 mmol/l in NIDDM, ~ 600 mmol/l in controls). FFA inhibited insulin-stimulated G_Rd , GS, or GLS similarly in patients with NIDDM and in normal control subjects, regardless of blood insulin and FFA levels. g , lipid; , saline. Adapted from Boden et al. (14,15). The conclusion that lipid infusion produced a transport/phosphorylation defect was supported by still another finding. Glycogen synthase activity was normal in muscle biopsies obtained 4 hours after lipid/heparin infusions, i.e., at a time when insulin stimulated glucose uptake was maximally inhibited (14). Since glycogen synthase is the rate limiting enzyme in the glycogen synthesis pathway, these findings suggested that flux through the glycogen synthesis pathway was intact after 4 hours of lipid/heparin infusion. This, however, changed during the ensuing 2 hours, i.e., after 4-6 hours of lipid/heparin infusion, when elevated plasma FFA caused a marked inhibition of muscle glycogen synthase activity associated with an increase in muscle glucose-6-phosphate concentration (14). Thus, elevated plasma levels of FFA produced at least two distinct biochemical defects 1) inhibition of insulin stimulated glucose transport and/or phosphorylation (after 3-4 hours of lipid/heparin infusion) and 2) inhibition of muscle glycogen synthase activity (after more than 4 hours of L/H infusion) "figure 6". Figure 6.* Defects of glucose utilization produced by FFA. The inhibition of carbohydrate oxidation (defect 1) was the earliest demonstrable defect. It developed during the initial 2 h of lipid infusion, but did not inhibit insulin-stimulated glucose uptake or glycolysis. The inhibition of glucose transport and/or phosphorylation (defect 2) developed after 3-4 h, while the inhibition of glycogen synthesis (defect 3) developed after 4-6 h of high plasma FFA. E.C., extracellular; I.C., intracellular. A third defect, namely a FFA mediated inhibition of CHO oxidation, which was first reported by Randle et al. in rat hearts (27) developed earlier than the other two defects (Figure 2). This defect, however, did not produce insulin resistance as glucose uptake was not impaired during the initial 3-4 hours of lipid/heparin infusion when CHO oxidation was already severely inhibited. Carbons which had entered the glycolytic pathway and could not oxidized because of the FFA produced increase in acetyl-CoA (13) and the ensuing inhibition of pyruvate dehydrogenase (28) were shunted into non-oxidative glycolysis (lactate/alanine production) (15). 4.4. Mechanisms The mechanisms responsible for the FFA induced defects in glucose transport and/or phosphorylation and glycogen synthesis are not known. Putative mechanisms include 1) activation of the hexosamine pathway, 2) inhibition of glucose transporter gene expression and 3) changes in cellular membrane fluidity. Marshall et al. have shown in cultured rat hepatocytes that prolonged hyperglycemia produced insulin resistance (glucose toxicity) by activating the hexosamine pathway (29,30). This pathway accounts for only 1-3% of glucose flux under normal conditions and results in the generation of several metabolites which are important substrates for glycoprotein and phospholipid synthesis. Recently, Rossetti=s laboratory has shown in rats, that fat induced insulin resistance was associated with accumulation of UDP-N-acetyl glucosamine, an end product of the hexosamine pathway, and that the same degree of insulin resistance could be reproduced by increasing UDP-N-acetyl hexosamine in skeletal muscle (31,32). Long and Pekala have shown that several long chain fatty acids decreased mRNA levels of the insulin responsive glucose transporter Glut 4 in fully differentiated 3T3-L1 cells by decreasing Glut 4 gene transcription and by destabilizing the Glut 4 message (33). Thus, FFA may cause insulin resistance by inhibiting Glut 4 gene expression in muscle. Lastly, FFA may induce changes in cell membrane fluidity. Insulin receptors are imbedded in the lipid bilayer of plasma membranes. There is some evidence to suggest that altering the fatty acid content of membranes can alter insulin receptor accessibility, insulin binding and action. For instance, increasing polyunsaturated fatty acid content has been found to increase membrane fluidity, insulin binding and action whereas decreasing their content had the opposite effect (34-37).

[Frontiers in Bioscience 3, d169-175, February 15, 1998]
Reprints
PubMed
CAVEAT LECTOR

FREE FATTY ACIDS (FFA), A LINK BETWEEN OBESITY AND INSULIN RESISTANCE

Guenther Boden, M.D.

Division of Endocrinology/Diabetes/Metabolism and the General Clinical Research Center, Temple University School of Medicine, Philadelphia, PA

Received 1/21/98 Accepted 1/31/98

4. FFA AND PERIPHERAL INSULIN RESISTANCE

4.1. Effect of FFA on Basal Glucose Uptake

When plasma FFA concentrations were increased acutely by IV infusion of a triglyceride emulsion (Liposyn II, 10% safflower and 10% soybean oil) and heparin (which enhances lipoprotein lipase activity), FFA levels rose 2-3 fold (to between 1.0 and 1.5 mM) but basal rates of glucose uptake did not change (12).

4.2 Effect of FFA on Insulin Stimulated Glucose Uptake

In healthy subjects, acute elevation of plasma FFA by IV lipid/heparin infusion inhibited total body glucose uptake dose dependently after a lag period of 3-4 hours "figure 1". Inhibition of CHO oxidation occurred ~ 2 h earlier than the inhibition of glucose uptake. The latter was reversed ~ 3 hours after discontinuation of lipid/heparin infusions (13) "figure 2". Other studies showed that the lipid induced inhibition of insulin stimulated glucose uptake was linear (14,15) "figure 3" and occurred similarly in healthy subjects and in patients with Type II diabetes (14-20). The long delay of 3-4 hours between the start of the insulin and lipid infusions and the development of significant insulin resistance was the most likely reason why the inhibitory effect of FFA on glucose uptake was not found in many previous studies (21-25).

Figure 1: Effect of plasma FFA on insulin-stimulated glucose uptake. Euglycemic-hyperinsulinemic (~ 420 pmol/l) clamping was performed in healthy volunteers for 6 h. High levels of plasma FFA were produced by the infusion of triglycerides (4.3 mmol/min) plus heparin (0.4 U A kg^-1 . min^-1) (D , n = 4); intermediate plasma FFA levels, by infusion of triglycerides without heparin (o, n = 4); and low FFA levels, by infusion of saline alone (·, n = 6). Data shown are mean ± SE. The inhibition of insulin-stimulated glucose uptake became statistically significant ~ 3.5 h after the start of the lipid infusion. *p < 0.05; **p < 0.01, comparing high with low FFA. Adapted from Boden et al. (14*).

Figure 2. Plasma FFA levels, and rates of CHO oxidation (COX) and glucose disappearance (G_Rd) in healthy men during lipid/heparin infusion (solid circles) and during lipid/heparin infusion from 0-120 min followed by saline infusion from 120-360 min (open circles) * p < 0.05, **p < 0.01, *** p < 0.005 comparing the 2 groups. Adapted from Boden et al. (13*).

The linear relationship between plasma FFA levels and insulin stimulated glucose uptake in healthy controls and patients with Type II diabetes suggested that for every 100 mM increase in plasma FFA, peripheral insulin sensitivity decreased by ~ 8% "figure 3". It needs to be emphasized, however, that the total range of insulin stimulated glucose uptake was approximately 2 fold higher in healthy controls than in patients with Type II diabetes (15) "figure 4". Thus, FFA could account for maximally 50% of peripheral insulin resistance in patients with Type II diabetes. Several potential problems could affect the interpretation of these results. First, the infused triglyceride emulsion (Liposyn II) contained a considerable amount of glycerol (2.5 grams/100 ml). We have shown, however, that this amount of glycerol did not affect glucose uptake (15). Second, the data showing that elevation of plasma FFA decreased peripheral glucose uptake in response to insulin were all produced in acute experiments. Their relevance relative to the long term effects of elevated plasma FFA, for instance during obesity could, therefore, be questioned. It was, however, shown that lowering of plasma FFA below basal values caused an increase in insulin stimulated glucose uptake (13) "figure 1". This indicated that basal FFA levels exerted long term inhibitory effects on peripheral glucose uptake.

Figure 3. Relationship between insulin stimulated glucose uptake and plasma FFA in 7 patients with NIDDM and in 6 non-diabetic controls. Data from controls were obtained during euglycemic (~ 5 mM)-hyperinsulinemic (~ 450 pM) clamping. Data from patients with NIDDM were obtained during hyperinsulinemic (~ 11 mM)-hyperinsulinemic clamping. Adapted from Boden et al. (14,15*).

Figure 4. Insulin-stimulated glucose uptake at comparable low plasma FFA (< 100 mmol/l) and euglycemia in diabetic and nondiabetic subjects. Shown are insulin-stimulated glucose uptakes before (g) and after ( ) 4 h of euglycemic (~ 4.8 mmol/l) hyperinsulinemic (~ 500 pmol/l) clamping in 7 patients with NIDDM and 6 nondiabetic control subjects. Preclamp glucose uptake could not be obtained in the diabetic patients because insulin was infused to lower their blood glucose concentrations into the normal range. Insulin-stimulated glucose uptake (G_Rd) was ~ 2 times higher in nondiabetic than in diabetic individuals (30 vs. 58 mmol/kg fat free mass, p < 0.01). Triglyceride plus heparin infusion (P) decreased insulin-stimulated glucose uptake by ~50% in diabetic and nondiabetic individuals. EU, euglycemia; FFM, fat free mass. Adapted from Boden et al. (14,15*).

4.3 Cellular Location of FFA Induced Defects

To determine the cellular location of the fat induced inhibition of insulin stimulated glucose uptake, glucose fluxes through all major pathways of intracellular glucose utilization were determined non-invasively. Rates of glucose uptake and glycolysis were estimated with 3-³H-glucose and glycogen synthesis was obtained by subtracting glycolysis from glucose uptake. The validity of these methods has been validated (26). CHO oxidation was determined by indirect calorimetry and non-oxidative glycolysis (lactate and alanine formation) by subtracting CHO oxidation from glycolysis. Using these methods, it was found that lipid/heparin infusion inhibited rates of glucose uptake, glycogen synthesis and glycolysis about equally (15). In normal controls, for instance, plasma FFA concentrations of ~ 600 mM inhibited insulin stimulated glucose uptake, glycogen synthesis and glycolysis all by ~ 50% while in patients with Type II diabetes, a higher plasma FFA concentration of ~ 1200 mM resulted in a ~ 90% inhibition "figure 5". These results suggested a FFA induced defect at the level of transport and/or phosphorylation (the methods used could not differentiate between these two possibilities). A primary defect at the level of glycogen synthesis or glycolysis, on the other hand, would have produced disproportionate reductions in the flux rates through these pathways.

Figure 5. Effects of elevated plasma FFA on rates of glucose uptake (G_Rd), glycogen synthesis (GS), and glycolysis (GLS) in 7 patients with NIDDM during hyperinsulinemic (~ 900 pmol/l) isoglycemic (~ 11 mmol/l) clamping and in 6 nondiabetic control subjects during euglycemic-hyperinsulinemic (~ 500 pmol/l) clamping. Total length of bars represent insulin-stimulated G_Rd , GS, or GLS, set as 100%. The darkly shaded parts of the bars represent insulin-stimulated G_Rd , GS, or GLS after 4 h of elevated plasma FFA (~ 1,200 mmol/l in NIDDM, ~ 600 mmol/l in controls). FFA inhibited insulin-stimulated G_Rd , GS, or GLS similarly in patients with NIDDM and in normal control subjects, regardless of blood insulin and FFA levels. g , lipid; , saline. Adapted from Boden et al. (14,15*).

The conclusion that lipid infusion produced a transport/phosphorylation defect was supported by still another finding. Glycogen synthase activity was normal in muscle biopsies obtained 4 hours after lipid/heparin infusions, i.e., at a time when insulin stimulated glucose uptake was maximally inhibited (14). Since glycogen synthase is the rate limiting enzyme in the glycogen synthesis pathway, these findings suggested that flux through the glycogen synthesis pathway was intact after 4 hours of lipid/heparin infusion. This, however, changed during the ensuing 2 hours, i.e., after 4-6 hours of lipid/heparin infusion, when elevated plasma FFA caused a marked inhibition of muscle glycogen synthase activity associated with an increase in muscle glucose-6-phosphate concentration (14). Thus, elevated plasma levels of FFA produced at least two distinct biochemical defects 1) inhibition of insulin stimulated glucose transport and/or phosphorylation (after 3-4 hours of lipid/heparin infusion) and 2) inhibition of muscle glycogen synthase activity (after more than 4 hours of L/H infusion) "figure 6".

Figure 6. Defects of glucose utilization produced by FFA. The inhibition of carbohydrate oxidation (defect 1) was the earliest demonstrable defect. It developed during the initial 2 h of lipid infusion, but did not inhibit insulin-stimulated glucose uptake or glycolysis. The inhibition of glucose transport and/or phosphorylation (defect 2) developed after 3-4 h, while the inhibition of glycogen synthesis (defect 3) developed after 4-6 h of high plasma FFA. E.C., extracellular; I.C., intracellular.

A third defect, namely a FFA mediated inhibition of CHO oxidation, which was first reported by Randle et al. in rat hearts (27) developed earlier than the other two defects (Figure 2). This defect, however, did not produce insulin resistance as glucose uptake was not impaired during the initial 3-4 hours of lipid/heparin infusion when CHO oxidation was already severely inhibited. Carbons which had entered the glycolytic pathway and could not oxidized because of the FFA produced increase in acetyl-CoA (13) and the ensuing inhibition of pyruvate dehydrogenase (28) were shunted into non-oxidative glycolysis (lactate/alanine production) (15).

4.4. Mechanisms

The mechanisms responsible for the FFA induced defects in glucose transport and/or phosphorylation and glycogen synthesis are not known. Putative mechanisms include 1) activation of the hexosamine pathway, 2) inhibition of glucose transporter gene expression and 3) changes in cellular membrane fluidity. Marshall et al. have shown in cultured rat hepatocytes that prolonged hyperglycemia produced insulin resistance (glucose toxicity) by activating the hexosamine pathway (29,30). This pathway accounts for only 1-3% of glucose flux under normal conditions and results in the generation of several metabolites which are important substrates for glycoprotein and phospholipid synthesis. Recently, Rossetti=s laboratory has shown in rats, that fat induced insulin resistance was associated with accumulation of UDP-N-acetyl glucosamine, an end product of the hexosamine pathway, and that the same degree of insulin resistance could be reproduced by increasing UDP-N-acetyl hexosamine in skeletal muscle (31,32).

Long and Pekala have shown that several long chain fatty acids decreased mRNA levels of the insulin responsive glucose transporter Glut 4 in fully differentiated 3T3-L1 cells by decreasing Glut 4 gene transcription and by destabilizing the Glut 4 message (33). Thus, FFA may cause insulin resistance by inhibiting Glut 4 gene expression in muscle.

Lastly, FFA may induce changes in cell membrane fluidity. Insulin receptors are imbedded in the lipid bilayer of plasma membranes. There is some evidence to suggest that altering the fatty acid content of membranes can alter insulin receptor accessibility, insulin binding and action. For instance, increasing polyunsaturated fatty acid content has been found to increase membrane fluidity, insulin binding and action whereas decreasing their content had the opposite effect (34-37).