WO2001010449A1

WO2001010449A1 - Method of maintaining vascular integrity using aicar (5-amino-4-imidazole riboside) and related compounds

Info

Publication number: WO2001010449A1
Application number: PCT/US2000/040607
Authority: WO
Inventors: Neil Ruderman; Yasuo Ido
Original assignee: Trustees Of Boston University
Priority date: 1999-08-09
Filing date: 2000-08-09
Publication date: 2001-02-15
Also published as: AU7629200A

Abstract

Long-term usage of AICAR (5-amino, 4-imidazole carboxamide riboside) has been found to decrease in mammals atherosclerosis and increase insulin sensitivity, resulting in benefits in diseases such as diabetes, hypertension, and gallstones. In addition, long-term usage of AICAR has an impact on amount of food consumed and results in reduced fat build-up and increase in muscle mass. AICAR appears useful to prevent vascular diseases associated with hyperglycemia, high plasma levels of free fatty acids (FFA) and insulin resistance by administering an agent that activates fatty acid oxidation. Thus, AICAR appears to be useful both for treatment and prevention of vascular diseases. Animal tests have shown that that chronic treatment with AICAR has not resulted in any noticeable toxic effects.

Description

TITLE OF THE INVENTION METHOD OF MAINTAINING VASCULAR INTEGRITY USING AICAR (5-AMINO-4-IMIDAZOLE RIBOSIDE)

AND RELATED COMPOUNDS

CROSS REFERENCE TO RELATED APPLICATIONS This application claims priority to U.S. Patent Application No. 60/147,923 filed August 9, 1999, which is incorporated by reference herein.

STATEMENT REGARDING FEDERALLY SPONSORED RESEARCH OR

DEVELOPMENT This project was partly funded by grant HL-55854-05 from the NIH.

BRIEF SUMMARY OF THE INVENTION Long-term usage of AICAR (5-amino, 4-imidazole carbox- amide riboside) has been found to produce metabolic and biological changes that should decrease atherosclerosis and increase insulin sensitivity, resulting in benefits in diseases such as diabetes, hypertension, and gallstones. In addition, long-term usage of AICAR has an impact on amount of food consumed and results in reduced fat build-up and increase in muscle mass. AICAR appears useful to prevent vascular diseases associated with hyperglycemia, high plasma levels of free fatty acids (FFA) and insulin resistance by virtue of the fact that this agent activates fatty acid oxidation. Animal tests have shown that chronic treatment with AICAR has not resulted in any noticeable toxic effects.

BRIEF DESCRIPTION OF THE DRAWINGS Figure 1 shows the Effects of AICAR on glucose induced apoptosis in HUVEC.

Figure 2 shows the mechanism by which AICAR inhibits apoptosis.

Figure 3 shows AMP-activated protein inase activity.

Figure 4 shows Acetyl Co-A carboxylase activity.

Figure 5 shows fatty acid oxidation in the presence and absence of AICAR.

Figure 6 shows the effect of AICAR administration (3 days/week) on food intake in rats on chow diet.

Figure 7 shows blood glucose following injection of AICAR.

Figure 8 shows the effects of AICAR on HUVEC metabolism after 16 hours incubation with 30 mM glucose.

Figure 9 shows the effects of 2-bromopalmitate and high levels of free fatty acid on apoptosis in HUVEC.

Table 1 shows the effects of AICAR incubation on metabolism after 2 hours with no carnitine. Table 1A shows ATP generation in HUVEC over 2 hours, with carnitine and prelabeling, including data for the second hour.

Table 2 shows the effect of 16 weeks of treatment with AICAR on body and organ weight.

Table 3 shows the acute effects of AICAR on plasma metabolites and hormones.

Table 4 shows the effect of 25 days of treatment with AICAR on plasma metabolites and hormones.

DETAILED DESCRIPTION OF THE INVENTION

AICAR (5-amino, 4-imidazole carboxamide riboside) when metabolized by cells yields a compound, ZMP which activates an AMP-activated protein kinase (AMPK) . We examined and report herein the useful effects of long-term usage of AICAR on vascular cells.

AICAR (also known as acadesine) , a purine nucleoside analog, has been known for many years to increase the generation of adenosine in ischemic myocardium and, for this reason, it has undergone trials to evaluate its effects on perioperative mycordial ischemia (Leung, 1994; Mangano, 1997) . AICAR also has been shown to activate AMPK in skeletal muscle and liver (Rev. by Ruderman, 1999) and by doing so it exerts a wide variety of metabolic effects on these tissues. In muscle, it reproduces many of the effects of exercise, including phosphorylation and inhibition of acetyl CoA carboxylase, and increases in fatty acid oxidation and glucose transport. In rat thymocytes, it has been shown to inhibit apoptosis (Stefanelli, 1998) and in cardiac endothelium, it can activate or inhibit nitric oxide synthase (Chen, 1999) . We have recently found that AICAR activates AMPK and inhibits acetyl CoA carboxylase in human umbilical vein endothelium. To our surprise, it diminished glucose uptake, glycolosis and ATP generation, but caused an apparent increase in the concentration of ATP. Thus AICAR either diminishes ATP use to an even greater extent than it diminishes ATP generation or it enhances other processes that generate ATP. As AMPK appears to be activated in response to ischemia and other stresses, we believe its activation reflects an effort of the cell to maintain its "energy" for processes required for its survival. In searching for ATP-consuming processes affected by AICAR in vascular cells (endothelium) , we have found that it inhibits non-oubain dependent Rb⁺ uptake (i.e., potassium uptake not due to Na⁺K⁺ ATPase) and apoptosis induced by hyperglycemia (Figure 1) . With respect to energy generation, we have found that AICAR markedly enhances fatty acid oxidation. (See Figure 5 and Table 1A. ) (Note: carnitine was added to several experiments to simulate the presence of carnitine in vivo. ) AMP-activated protein kinase (AMPK) is a cytoplasmic enzyme that has been shown to exist in both the liver and skeletal muscle (Abu-Elheiga, 1997; Ha, 1994; Hardie, 1989; Hardie, 1997) . As its name indicates, AMPK is activated by increasing levels of AMP or the ratio of AMP to ATP in the cell. AMP levels rise in the cell as ATP is hydrolyzed to ADP and P,-. Two molecules of ADP, through the action of myokinase, also known as adenylate kinase, produce one molecule of ATP and one molecule of AMP. In addition to its activation by AMP, AMPK is activated through phosphorylation by an upstream kinase called AMPK kinase (AMPKK) (Hardie, 1997). AMP also allosterically activates AMPKK (Hardie, 1997) . Phosphorylation of AMPK by AMPKK makes it a poor substrate for proteases. All these factors combined together make AMPK very sensitive to minimal fluctuations in cellular AMP levels (Hardie, 1997) . 7ΛMPK has several known substrates, specifically enzymes which it can phosphorylate and modulate. In the liver, 7ΛMPK has been shown to phosphorylate hydroxymethyl glutaryl CoA

(HMGCoA) reductase (Hardie, 1989) and acetyl CoA carboxylasse (ACC) (Abu-Elheiga, 1997; Ha, 1994; Hardie, 1989; Hardie, 1997), inhibiting both their actions. Reducing HMGCoA reductase activity inhibits cholesterol synthesis (Hardie, 1989) , and reducing ACC activity decreases fatty acid synthesis and malonyl CoA generation. In skeletal muscle, AMPK also phosphorylates and inhibits ACC (Hardie, 1997; Kim, 1997; Kim, 1998). In addition, it has been shown to increase glucose transport into the muscle (Merrill, 1997).

ACC is the enzyme responsible for the first committed step in fatty acid synthesis, the carboxylation of acetyl CoA in the cytosol of a cell to produce malonyl CoA. Citrate is both a precursor of acetyl CoA and an allosteric activator of ACC (Allred, 1997; Kim, 1989) . Malonyl CoA, an intermediate of fatty acid synthesis (Wakil, 1983) inhibits carnitine palmitoyl transferase I (CPTI), the enzyme that regulates the uptake of long chain fatty acid CoA (LCFA CoA) into the mitochondria where they undergo oxidation (McGarry, 1995) . Thus, by sensing an increase in total AMP levels, which parallels a need for energy, AMPK allows the cell to switch from storing fatty acids as triglyceride (i.e., esterified lipid) to oxidizing it to provide free energy for its biological needs. In liver, this appears to be accomplished by decreasing the storage of cholesterol and fatty acids (i.e., by decreasing energy use) as well as by allowing fatty acid oxidation to occur at a higher rate and thus provide more ATP. To stimulate AMPK experimentally, a compound called AICAR (5-aminoimidizole-4-carboxamide riboside) is used. AICAR enters the cell where it is phosphorylated to become ZMP, an AMP analog (Hardie, 1997) . ZMP activates both AMPK and AMPKK. In turn, AMPKK phosphorylates AMPK and activates it even further (Hardie, 1997) .

As a result of previous studies demonstrating the presence of AMPK in muscle, liver and other tissues, it is of interest to further investigate its existence in the vasculature, and its potential role in modulating vascular metabolism and function. The present studies have been conducted in human umbilical vein endothelial cells (HUVEC) . We show that AMPK is indeed present in endothelial cells and its activity is enhanced by the administration of AICAR. In addition, we demonstrate the presence of ACC in these cells and the modulation of its activity with the rise of AMPK. We also show the metabolic changes such as glucose uptake, glycolysis, glucose oxidation and fatty acid oxidation, associated with increase in AMPK activity. Changes in rubidium uptake are also included. Some work has been done to evaluate the short-term usage of AICAR. US Patent No. 5,443,836 discloses that usage has resulted in an increase in the local concentration of adenosine, which may benefit patients with a wide variety of disorders associated with decreased blood flow (ischemia) . These disorders include stroke, heart attack and adverse effects associated with ischemia of the liver, bowel and, by inference, other organs. All of the studies relative to this use of the agent were performed in humans or experimental animals for relatively short periods of time (usually less than 80 hrs) . US Patent No. 5,658,889 discloses that the short-term usage of AICAR in very high doses (500 mg/kg/twice daily) lowers blood glucose levels in control and diabetic rats. They also carried out studies in which diabetic rats were treated with such a regimen for 23 days with an apparent decrease in the severity of the diabetes as judged by lower blood glucose levels and decreased polyuria (i.e., a decrease in the large urine volume) . Based on this work, they propose that AICA-riboside could be useful in treating pathological conditions including insulin deficiency, insulin resistance, diabetes and syndrome X (also now referred to as the insulin resistance syndrome or the metabolic syndrome) . No assessment of insulin resistance was presented in their patent, however, and the basis for their conclusions about insulin resistance is unclear. US Patent No. 5,234,404 discloses the diagnosis, evaluation and treatment of coronary artery disease by exercise simulation using closed loop drug delivery of an exercise simulating agent beta agonist.

It has been disclosed (R.A. Cohen, N. B. Ruderman, Y. Ido, unpublished data) that, in rats treated with AICAR for 106 days, aortic rings stripped of endothelium responded to said treatment. In brief, we found that the vasodilatory response of the rings to nitric oxide was twice as great in the AICAR-treated rats as in control rats. Since nitric oxide is thought to protect the arterial wall against athero-sclerosis and thrombosis, and it can diminish apoptosis, these findings suggest another mechanism by which AICAR could protect against vascular disease.

Based on our observation of diminished apoptosis in endothelial cells, we believe that treatment with AICAR, or a derivative thereof, will be useful in particular in patients with diabetes mellitus, especially non-insulin dependent diabetes mellitus (NIDDM) . Treatment with AICAR should diminish damage to the endothelium caused by hyperglycemia, thereby reducing or preventing coronary disease. By virtue of its effects on glucose and fatty acid metabolism in pericytes, it should also be useful in preventing and treating the microvascular complications of diabetes (such as blindness, retinopathy and nephropathy) .

Furthermore, AICAR should also prove a useful tool as a chronically administered therapeutic agent in a wide array of situations in which endothelial cell integrity is compromised by stress - e.g., hyperglycemia, high plasma free fatty acid (FFA) levels and possibly by ischemia and inflammation. AICAR is poorly absorbed from the gastrointestinal tract (ca. 5%); therefore, it has been administered parentally in vivo. Thus, in addition to assessing the physiological efficacy of AICAR, efforts will be undertaken by methods well known to skilled chemical pharmacists to develop derivatives of AICAR or other AMPK activators that are better absorbed. The proposed mechanism of AICAR is believed to be the following. By activating AMPK, AICAR inhibits acetyl CoA carboxylase (ACC) . Inhibition of ACC in turn results in a decrease in the concentration of malonyl CoA, an inhibitor of carnitine palmitoyl transferase, an enzyme that controls fatty acid oxidation by regulating the transfer of long chain fatty acyl CoA (LCFA CoA) into mitochondria. We have shown in cultured human umbilical vein endothelial cells that treatment with AICAR activates fatty acid oxidation, diminishes fatty acid esterification and inhibits the programmed cell death (apoptosis) caused by high concentrations of glucose and/or fatty acids. We have also shown that in the latter situation the ability of insulin to activate the enzyme Akt/PKB, which by itself inhibits apoptosis, is depressed. In addition we have demonstrated that AICAR overcomes this abnormality in insulin action. Others (Chen 1999) have shown that activation of AMPK is associated with the phosphorylation of a specific serine residue on the endothelial form of nitric oxide synthase (eNOS) and that, dependent on the condition, this can result in an increase in nitric oxide synthesis. Interestingly, insulin can also increase NO synthesis, and this is associated with the phosphorylation of eNOS on the same serine residue phosphorylated by Akt/PKB. (See Figure 2 for proposed mechanism by which AICAR inhibits apoptosis.)

We have demonstrated the presence of both AMPK and ACC in the vasculature. We show that AICAR increases AMPK, leading to a decrease in ACC and thus an increase in fatty acid oxidation, as seen in both liver and skeletal muscle. In addition, we show a decrease in glycolysis and increase in glucose and fatty acid oxidation and possibly total levels of ATP. This increase or maintenance of cellular ATP was not associated with a decrease in protein synthesis or Na/K pump activity, although it was shown to decrease flux through an as yet unidentified K⁺ channel. The major effect of AICAR on ATP maintenance appeared to be due to a large increase in fatty acid oxidation; indeed, we found that nearly 60% of the ATP generated by HUVEC incubated with AICAR was accounted for by fatty acid oxidation (see Table 1A) . Experiments have indicated that AICAR can be administered for upwards of 3 months without obvious toxicity. Our preliminary studies indicate that rodents can tolerate AICAR at a dose of 250 mg/kg administered subcutaneously for upwards of 3 months without evidence of gross toxicity, suggesting it can be used chronically.

The data show that AICAR substantially diminishes intra-abdominal fat without diminishing the mass of other organs; indeed, if anything, our preliminary data suggest it increases muscle mass. In addition, chronic treatment with AICAR diminished plasma leptin and insulin levels in keeping with decreases in adiposity, and it decreased plasma triglycerides and possibly cholesterol. All of these findings suggest that AICAR chronically increases insulin sensitivity and decreases adiposity and plasma lipids—all of which should decrease predisposition to atherosclerosis. The improvement in insulin sensitivity should also decrease the risk of other diseases associated with the insulin resistance syndrome (e.g., diabetes, hypertension, gallstones) in humans. We are proposing that the chronic (indefinite time) use of such agents can prevent or retard the development of various types of vascular disease associated with metabolic abnormalities. These would include atherosclerotic vascular disease and, in particular, atherosclerosis associated with diabetes and insulin resistance and diabetic microvascular disease.

We believe that similar benefits would be found in other mammals. The dosage levels used in rats (above) correspond to approximately 50 mg/kg human body weight/day or about 3-4 grams dosage per human/day. The following examples are intended to further illustrate, but not limit, the invention.

Example 1: Studies with Cultured HUVEC Experimental Procedures Material. Cell culture materials were purchased from Clonetics (San Diego, CA) . All radioactive chemicals were obtained from NEN Life Science Products (Boston, MA) . All reagents were purchased from Sigma (St. Louis, MO), unless otherwise indicated. Cell Culture. Human umbilical vein endothelial cells (HUVEC) were plated in T75 flasks for oxidation studies and 100 mm plates for AMP-kinase and ACC assays. Cells were grown in a 37 °C, 5% C0₂ incubator. Incubation with AICAR. Cells were washed with warm PBS, pre-incubated with Earle' s Balanced Salt (1.8 mM CaCl₂, 0.8 mM MgS0₄, 5.3 mM KCl, 116 mM NaCl, 1 mM NaH₂PO₄)/20 mM HEPES solution pH 7.4 for 30 minutes, then incubated with Earle' s HEPES solution ±AICAR for the indicated time and concentration at 37 °C, 5% Co₂. Protein Determination. Protein concentration in the samples was determined by using BCA assay from Pierce (Rockford, IL) .

AMP-Kinase Assay. AMPK activity was determined by measuring phosphorylated SAMS peptide, (QCB, Inc.) a synthetic peptide substrate for AMPK (Witters, 1992). Cells in 100 mm dishes were incubated with 3 ml of medium ±AICAR at the indicated concentrations. After incubation, the cells were washed with cold PBS and lysed with 1 ml of ice cold lysis buffer (20 mM Tris HCl pH 7.4 at 4°C, 50 mM NaCl, 50 mM NaF, 30 mM NaPPi, 250 mM sucrose, 10 μN ZnCl₂, 100 mM Na-vanadate, 2 mM DTT, 100 μg/ml PMSF, 1 μg/ml pepstatin A, 1 μg/ml leupeptin and 0.4 mg/ml digitonin) . The cells were scraped, homogenized then centrifuged at 14,000 xg for 30 minutes at 4°C. The supernatants were concentrated with ammonium sulfate with 35% saturation (Hutber, 1997). Samples were kept on ice for 1 hour, then centrifuged at 14,000 xg for 30 minutes. The pellet was resuspended with 100 μl buffer

(lysis buffered with no digitonin) . The assay mixture consisted of 10 μl sample and 40 mM K-KEPES, 80 mM NaCl, 1 mM DTT, 0.2 mM SAMS peptide, ±0.2 mM 5'-AMPm 0.2 mM ³²P-ATP. After washing with 75 mM phosphoric acid twice for 5 minutes and once for 15 minutes, then with phosphate buffer pH 7.5 for 2 minutes, P81 paper were counted for ³²P to determine phosphorylation of SAMS peptide.

Acetyl CoA Carboxylase . Acetyl CoA carboxylase activity was determined by the amount of malonyl CoA produced through fixation of ¹⁴Co₂ according to the Lowenstein method. After incubation with AICAR, the cells were lysed with 0.1% NP-40 in Buffer A (50 mM Tris, pH 7.5, 1 mM DTT, 1 mM EDTA, 5 μM PMSF, 5 μM Aprotinin, 5μM leupeptin, 5 μM Pepstatin 20 mM β- glycerophosphate, 20 mM NaF, 2 mM NaPPi and 1 mM Na- vanadate) . The cells were scraped, homogenized with a Dounce homogenizer (20 strokes) then centrifuged for 15 minutes at 14,000 xg. A 50 μl sample was incubated with buffers C (same as buffer B) and D' or in buffers C and D. Buffers C and D' contain no citrate while C and D contain citrate. Buffers C and C are made up of Buffer A plus 20 mM MgC12 and 1 mg/ml bovine serum albumin (fatty acid free) . At this point, 10 mM citrate is added to buffer C. Buffers D and D' were made up by adding to buffers C and C , respectively, 7 mM ATP, 0.25 mM acetyl CoA and 25 mM of high specific activity NaH¹⁴C03. The reaction mixture consisting of a 50 μl of sample, 50 μl buffer C or C and 100 μl of D or D' were incubated for 12 minutes at 37 °C. The reaction was stopped with 15% PCA and the mixture was transferred into scintillation vials were all the NaH¹⁴C03 was evaporated in an overnight incubation at 60 °C. The remaining ¹C, which is mainly malonyl CoA, was counted. Fatty Acid Oxidation. 3H- (U) palmitic acid (18 μCi/ml) was incubated with the cells and ³H20 as collected (Wanders, 1995) . After incubation, 1 ml of the media was transferred to test tubes and acidified (pH<2) with 10% TCA. The samples were then centrifuged at 2,000 xg for 5 minutes at 4°C. The supernatant was then neutralized with 6N NaOH. To separate the labeled water from palmitate, the samples were passed through a AG1-X8-OH form column from BioRad

(Hercules, CA) . The columns were washed twice with water and the elutant counted for ³H.

Glucose Oxidation. ¹C- (U) glucose (lOμCi/ml media) was used to measure glucose oxidation by collecting C0₂ produced. After incubation, 1 ml of the media was transferred to test tubes, capped with a rubber stopper to which was attached a well containing 300 μl IN NaOH. 200 μl of 10% PCA were injected into the media through the stopper to liberate C0₂ which was collected in the well. The content of the well was counted for ¹⁴C. Control experiments were performed by incubating T75 flasks with a ¹⁴C-labeled bicarbonate solution, treated in the same way as in the glucose oxidation experiment, to confirm that the C0₂ was retrieved quantitatively. Glucose Measurement in the Media. To determine glucose in media before and after incubation, Sigma Diagnostics Glucose HK kit was used where glucose is converted to 6- phosphogluconate and NADH causing an increase in absorbance at 340 nm directly proportional to glucose amount in the sample.

Lactate measurement. To measure lactate production in the cells, 1 ml of media was first acidified with 50 μl of 70% PCA, then neutralized with (2 M KHC0=3=, 2 MKOH) to reach a pH between 5 and 6. A 50 μl sample was incubated in a reaction mixture consisting of 0.19 M Hydrazine (10 mM HCl), 2 mM NAD⁺ and 20 units/ml of lactate dehydrogenase. The absorbance was read at 340 nm every 5 minutes for about 30 minutes when readings reached a plateau (Lowry, 1993) . Pyruvate measurement. Samples were incubated with 50 mM phosphate buffer pH 7.0, 0.1 mM NADH and 0.1 units/ml lactate dehydrogenase. The samples were read before and 4 minutes after addition of enzyme (Lowry, 1993) . ATP measurement. A bioluminometric method was used to determine ATP levels. A 10 μl sample is added to 100 μl of imidazole reagent (Imidazole-HCl, pH 7.0, (30 mM imidazole base, 20 mM imidazole-HCl) , 1 mM MgCl₂, 60 μM ADP, 0.02% bovine serum albumin and 10 nM AP5A) . A 10 μl sample of that is added to 200 μl of luciferin-luciferase reagent (glycylglycine buffer, pH 8.1 (24 mM glycylglycine base, 25 mM glycylglycine-HCl) , 2mM EGTA, 2 mM MgCl₂, 2 mM DTT, 0.04% bovine serum albumin, 10 nM AP5A, 18 μM luciferin, 0.05 units/ml luciferase) . The samples were read in a luminescence biometer (Lowry, 1993) .

Rubidium Uptake Assay. To measure Na/K pump uptake activity, 86-Rubidium, which behaves like potassium ions, was used. Cells were incubated with 0.5 μCi/ml Rb for 10 minutes then washed 4 times with ice cold 100 mM MgCl₂. The cells were then lysed with 0.1 N NaOH and 0.1% NP-40 and counted for total Rb uptake. To inhibit the Na/K pump, 0.2 mM ouabain was added to the media for 10 minutes. In order to demonstrate the presence of AMP-dependent protein kinase in HUVEC, confluent cells were incubated with varying concentrations of AICAR for 30 minutes (Figure 3A) . The enzyme was assayed in a reaction mixture containing either no AMP or 0.2 mM 5' -AMP. The difference in the activity is the AMP-activated kinase activity. The kinase activity increased from 6.7 ±0.2 pmol/min/mg protein at 0 MM

AICAR incubation, to 15.8 ±0.4 at 0.2 mM (P=0.05), 22.5 ±0.3 at 0.5 mM (P=0.05), and 30.9 ±0.2 at 2 mM AICAR incubation

(P=0.002). With increasing AICAR concentration, there was a significant increase in AMPK activity as expected. This indicates the presence of AMPK and AMPKK in HUVEC. Incubation of cells with 2 mM AICAR showed an increase in AMPK activity as compared to control. The change was seen at 30 minutes and persisted for at least 120 minutes (Figure 3B) . The kinase activity increased from 7.3 ±1.8 at 0 minutes to 30.8 at 30 minutes (P=0.01), 31.5 at 60 minutes (P=0.01) and 32.3 pmol/min/mg protein at 120 minutes (P=0.01) .

As previously mentioned, AMPK increases fatty acid oxidation in both liver and skeletal muscle by phosphorylation and inhibition of acetyl CoA carboxylase. We wanted to test if this effect of AMPK is also present in endothelial cells. Upon incubation with AICAR, ACC activity decreased. A significant decrease was seen within the first 30 minutes of incubation (Figure 4). ACC activity went down from 163.1 ±27.0 pmol/min/mg protein at 0 minutes to 61.1 ±21.1 at 30 minutes (P=0.004), 73.0 ±12.4 at 60 minutes and 51.2 ±5.0 at 120 minutes (P=0.015). As previously described, ACC activation increases malonyl CoA levels and this in turn inhibits carnitine palmitoyl transferase I (CPT1) , the enzyme responsible for the uptake of long chain fatty acyl CoA into the mitochondria where they are oxidized. If ACC activity goes down, as seen in Figure 4, one would expect the concentration of malonyl CoA to decrease and fatty acid oxidation to go up. In the absence of carnitine (Table 1) , fatty acid oxidation did not show any changes. However, when 50 μM carnitine was added, fatty acid oxidation increased significantly (Figure 5) . Fatty acid oxidation was measured by total accumulation of water production. A significant increase was only seen after a 2- hour incubation (P=0.002). In addition, the increase in fatty acid oxidation rose exponentially after 60 minutes of incubation for both + and -AICAR. (See Table 1A. )

Glucose and fatty acid oxidation, glucose uptake and lactate and pyruvate release were determined in the absence and presence of 50 μM carnitine. The presence of carnitine had an effect only on fatty acid oxidation. Changes in glucose uptake and oxidation and lactate and pyruvate release caused by AICAR were the same in both the presence (50 μM) and absence of carnitine.

It has been demonstrated, in contracting muscle and muscle incubated with AICAR, that glucose uptake increases as AMPK activity increases, thereby providing more fuel for the muscle cell (Merrill, 1997), Merrill, 1998). To see if this also occurred in endothelial cells, glucose uptake was measured (Table 1) . The results were the opposite of those seen in the muscle, in that glucose uptake was decreased during the AICAR incubation.

In addition, lactate and pyruvate production were measured to see if the rate of glycolysis had changed. Pyruvate production showed a slight, but significant decrease whereas lactate production showed an even more obvious decrease (Table 1) .

Despite the decrease in glycolysis (as evident from the decrease in lactate and pyruvate production) , the levels of ATP measured by bioluminescence in HUVEC after AICAR incubation, if anything, rose. This was not attributable to an increase in glucose oxidation. Glucose oxidation was measured by determining labeled C0₂ production when the cells are incubated with ¹C- (U) glucose. After two hours, glucose oxidation increased significantly; however, the amount of ATP generated by glucose oxidation was small (See Tables 1 and 1A) . The retrieval experiments with labeled bicarbonate showed about 82% retrieval of bicarbonate from the media.

Despite the maintenance or even increase in ATP levels in HUVEC incubated with AICAR, the calculated rate of ATP generation was significantly diminished (See Table 1) . Later studies revealed that this was because we underestimated the rate of fatty acid oxidation. Prior studies in skeletal muscle had revealed that radioactive fatty acids added to an incubation or perfusion medium or even to blood, first have to mix with intracellular lipids before they are oxidized. As a result their rate of oxidation can be greatly underestimated unless the tissue is prelabeled with fatty acid. We found that the same phenomenon also occurs in HUVEC (See Figure 5 and Table 1A) . Thus, when the HUVEC were pre-incubated with radioactive palmitate for 24 hours prior to ¹C0₂ collection (to label the intracellular lipid pools) , the measured rate of fatty acid oxidation was increased by 5 fold during the second hour of a two-hour incubation with radioactive palmitate added to the medium. (See Table 1A. ) During this time period, calculated ATP production in cells exposed to AICAR was similar to that of cells incubated without AICAR; however, fatty acid oxidation accounted for 60% of the ATP generated vs. only 8% in cells that were neither pre- incubated with radioactive palmitate or treated with AICAR. (See Table 1A. )

We also examined the possibility that ATP utilization decreased by AICAR. One of the major ATP users, the Na/K pump, was investigated. The activity of this pump was tested by measuring the uptake of radioactive rubidium in the absence and presence of ouabain, an Na/K pump inhibitor. The difference in uptake, which is referred to as ouabain- sensitive Rb⁺ uptake, is the Na/K pump activity. In this experiment, the cells were incubated with 2 mM AICAR for 2 hours. Total rubidium uptake decreased significantly from 12.9 ±1.5 pmol/min/μg protein to 6.0 ±0.3 pmol/min/μg protein at 0 and 2 mM AICAR, respectively (P=0.001, n=6) . When the ouabain sensitive component was calculated, no difference was seen between the two incubations, (2.11 ±0.9 and 3.3 ±0.9 pmol/min/μg protein) thus leading us to believe that no change in Na/K pump activity had occurred. On the other hand, total rubidium uptake was significantly decreased, suggesting that AICAR diminishes the activity of an as yet unidentified K⁺ channel.

Analysis of experimental results

AMP-activated protein kinase has been shown to exist in both liver and skeletal muscle. Its presence in the vasculature, however, has only been mentioned recently, but not directly shown (Chen, 1999) . In this experiment, we specifically demonstrate the presence of AMPK in endothelial cells and its effects on metabolism.

As previously described, AMPK activity can be induced experimentally by incubation of tissues with the drug AICAR. AICAR enters the cells and gets phosphorylated to ZMP, an AMP analogue. ZMP activates both AMPK and the upstream kinase AMPKK which, in turn, activates AMPK through phosphorylation .

Incubation of the HUVEC with different AICAR concentrations shows that AMPK activity increased significantly. Within the first 30 minutes, AICAR was able to induce AMPK. The activity of AMPK in the endothelial cells is lower than that seen in the liver and skeletal muscle, but the activation due to the presence of AMP is clear, demonstrating the presence of AMP-activated protein kinase. The AMPK activity remained high with a longer incubation time up to 120 minutes. The 2 mM AICAR was the concentration of choice for the remaining experiments due to its obvious activation of AMPK and its low toxicity to the cells .

One of the roles of AMPK is the sensing of AMP vs. ATP levels in the cell. Activation of AMPK in response to an increase in the AMP/ATP ratio allows the cell to modify its metabolic activities to provide energy when needed. As previously described in skeletal muscle and liver, AMPK phosphorylates acetyl CoA carboxylase, inhibiting its activity, thus reducing the levels of its substrate, malonyl CoA. ACC is the first committed step in fatty acid synthesis and thus energy storage. In addition, the decrease in malonyl CoA levels relieves the inhibition of CPT1, thus allowing long chain fatty acids to be taken up into the mitochondria where they are oxidized. Thus, modulating the activity of ACC through AMPK not only decreases energy storage but also allows for energy production.

The results described in other tissues are seen in the endothelial cells. Figure 4 demonstrates both the presence of ACC in the endothelial cells as well as its inhibition upon incubation with AICAR, i.e. activation of AMPK. The presence of ACC is shown by the substantial increase in enzyme activity when 10 mM citrate is present vs. 0 mM citrate. Citrate allosterically activates ACC, but not other enzymes, such as propionyl CoA carboxylase and pyruvate carboxylase, which could also use HC0₃ ^~ as substrate. Thus, the increase in radioactive HC0₃ ^" use in the presence of citrate is a reflection of ACC activity only. Similar to AMPK, the activity of ACC is lower in endothelial cells than in skeletal muscle and liver. Again, within 30 minutes of incubation in 2 mM AICAR, changes in ACC activity are seen, and remain low for 2 hours.

Decreases in ACC activity when AMPK is activated inhibit fatty acid synthesis in liver and some other tissues. However, endothelial cells have little ability, if any, to synthesize fatty acids. Thus this decrease in ACC must play another role. Indeed, when fatty acid oxidation in these cells was measured in the presence of carnitine, a significant increase was seen after 2 hours of incubation with 2 mM AICAR. (See Figure 5 and Table 1A. )

Despite the activation of AMPK and the deactivation of ACC within the first 30 minutes, no change in fatty acid oxidation was seen after 30- or 60-minutes of incubation with AICAR, although it was clearly observed at 120 minutes. This could occur if the decrease in malonyl CoA levels caused by ACC inhibition lagged behind the changes in DMPK and ACC. It could also be related to the slow equilibration of the fatty acids added to the medium with those present within the cells. In this context, it has been hypothesized that there are internal pools of fatty acids which the cell uses for β-oxidation and triglyceride synthesis. Exogenous fatty acids have to first be incorporated into these pools before they can be used by the cell. If true, this could explain in part the reason why fatty acid oxidation in HUVEC, assessed by the oxidation of exogenous fatty acids increased dramatically only after 60 minutes of incubation with AICAR.

Changes in fatty acid oxidation were not seen when carnitine was absent from the incubation media and the absolute rate of fatty acid oxidation was substantially lower. (See Tables 1 and 1A. ) Previous studies have shown that the addition of carnitine is necessary for the uptake of long chain fatty acids into the mitochondria in cultured cells incubated in media devoid of carnitine (Hulsmann, 1992, Hulsmann, 1988) . So far, we have found that AMPK activation leads to increased energy production by increasing fatty acid oxidation. We have also found that carnitine is necessary for this to occur, but that carniotine has no independent effect on the other actions of AICAR (e.g., glycolysis) that we have evaluated. / In contracting muscle, AMPK activity rises (Vavvas, 1997) . It has been shown that this as well as incubation .with AICAR, which increases AMPK activity, leads to an increase in glucose uptake (Merrill, 1997, Merrill, 1998). Contrary to what is seen in muscle, glucose uptake was reduced in HUVEC. In addition, incubation of endothelial cells with 2 mM AICAR showed a significant decrease in both lactate and pyruvate production after 2 hours. This decrease in glycolysis and glucose uptake is contrary to what is expected when AMPK is activated under circumstance where increased energy production is desirable. Such an inhibition of glycolysis upon incubation in AICAR (Vincent, 1992) has also been observed in isolated rat hepatocyte. However, they indicated a decrease in phosphorylation of glucose by glucokinase and a diminished 6-phosphofructo-l- kinase activity. Considering that endothelial cells have mainly insulin independent glucose transporters, it is possible that the AICAR-induced inhibition of glucose uptake occurs by a similar mechanism.

In contrast to glucose uptake in glycolysis, glucose oxidation showed a significant increase with AICAR incubation. This increase in glucose oxidation, however, does not account for the decrease in lactate/pyruvate release. Also, the ATP generated by glucose oxidation was much lower than that due to the fatty acid oxidation. Total ATP levels in the cells appeared to significantly increase after incubation with AICAR. AICAR itself did not affect the readings during the ATP assay; however the possibility that some of the apparent increase in cellular ATP was due to the ZMP metabolite, ZTP has not been ruled out.

Calculations of total energy production, from glycolysis, fatty acid and glucose oxidation, initially suggested that the production of ATP was lower in AICAR incubated cells, despite the significant increases in fatty acid oxidation. As shown in Table 1A, this is in large part due to the fact that fatty acid oxidation is grossly underestimated when the cells are not incubated with carnitine and prelabeled with radioactive fatty acid. We also examined the possibility that a decrease in energy use occurred in AICAR-treated cells. There are many ATP utilizers in cells and one or a combination of a few of them could be candidates for AMPK phosphorylation and inhibition directly or through other downstream mechanisms. One of the users is the Na/K pump, which could account for a large portion of ATP utilization. To test this, rubidium was used. It is a molecule that behaves like K+ and uses its pumps and channels to cross the membrane. Incubation of cells with 2 mM AICAR caused no change in Na/K pump activity, the ouabain-sensitive component of K+ uptake. However, ouabain insensitive rubidium uptake component decreased significantly. Since rubidium behaves like K+, it is possible that other K+ channels close with AICAR incubation. There are many types of K+ channels in cells and some of them are ATP-dependent . Further investigations are required to determine which of these channels is inhibited by AICAR and whether this effect is due to activation of AMPK or to an AMPK-independent effect of AICAR.

Example 2 :

Long-term administration of AICAR

AICAR was administered subcutaneously to rats at the indicated doses for periods ranging from 2 hours to 106 days. It was found that food intake was diminished by approximately 20% over the 24 hours following AICAR administration, but the rats seemed to compensate by eating more the following day. (See Figure 6) Total food intake was diminished by only 3% during the long-term studies.

At post-mortem evaluation, the mass of intra-abdominal (retroperitoneal, mesenteric, epididymal) fat deposits was diminished by 20-40% and muscle triglycerides by approximately 20%. No differences were observed in the weights of heart, liver or other organs in AICAR treated rats at 25 days. However, preliminary data suggests an increase in muscle mass at 106 days. (See Table 2.)

Others have found that AICAR moderately diminishes blood glucose 2-6 hours after its administration (Figure 7) . At 2 hours, we found that high doses of AICAR increased blood lactate and decreased plasma free fatty acids. (See Table 3.) AICAR at all doses studied did not significantly alter plasma triglycerides or leptin (Table 3) .

In contrast, in rats chronically treated with AICAR (25 days) , plasma leptin was significantly diminished (24 hours after last injection) in keeping with diminished adiposity. Muscle triglyceride was diminished by 25% and plasma triglyceride by 40%. (See Table 4 for plasma data.) Preliminary data (not shown) suggest that insulin sensitivity is increased in these rats.

Example 3:

Effect of AICAR on pericytes

Human and bovine retinal pericytes were treated with AICAR. It was found that treatment with AICAR reduced apoptosis in cells exposed to high fatty acid levels together with glucose. Thus, based on our hypothesis that treatment with AICAR should reduce blindness, retinopathy and other pathological changes in the eye in patients suffering from diabetes. Material. Cell culture materials were purchased from Clonetics (San Diego, CA) . All radioactive chemicals were obtained from NEN Life Science Products (Boston, MA) . All reagents were purchased from Sigma (St. Louis, MO), unless otherwise indicated. Cell Culture. Human retinal pericytes were plated in 6 well plates. Cells were grown in a 37 °C, 5% C0₂ incubator with SmBM medium.

Apoptosis induction and diacylglycerol (DAG) , ceramide measurement. Apoptosis was induced by incubating the cells with Medium 199 + 5% FBS medium containing 0 - 0.5 mM palmitic acid for 3 days. Apoptotic cells were determined by TUNEL staining. Ceramide and DAG levels were measured by diacylgycerol kinase method using 32P-ATP. To determine the effects of AICAR, 1 mM AICAR was added for these 3 days period. Increased fatty acid levels promoted apoptosis in human and bovine retinal pericytes in a dose dependent fashion (0.5% with 0.1 mM palmitate vs 27% with 0.5 mM palmitate). The apoptotic rate was further increased by incubating the cells with high glucose. ( with 0.2 mM palmitate, 4% of the cells were apoptotic in 5 mM glucose vs. 7% of the cells in 20 mM glucose) . This increased apoptosis was accompanied by increased intracellular ceramide levels (3 nmol/mg protein with 0.1 mM palmiate vs 10 nmol/mg protein with 0.5 mM palmitate) . Incubation with 1 mM AICAR decreased apoptosis by 50% (27±4 % with 0.5 mM palmitate vs 10±2% with 0.5 mM palmitate+1 mM AICAR) . AICAR also decreased ceramide levels by 60% (9+2 nmol/mg protein with 0.5 mM palmitate vs 3.8+1 nmol/mg protein with 0.5 mM palmitate + 1 mM AICAR) in human retinal pericytes and DAG levels by 50% (47±3 nmol/mg protein with 0.5 mM palmitate vs. 24±5 nmol/mg protein with 0.5 mM palmitate + lmM AICAR) .

Assessment of the Results The results are consistent with metabolic effects of AICAR depicted in Figure 2. They show that increased glucose levels enhance fatty-acid induced apoptosis possibly by enhancing free fatty acid esterification (DAG and ceramide), and that AICAR prevents these effects.

Example 4 : Additional Studies in HUVEC

Effects of AICAR on mitochondrial membrane potential, free acid oxidation and free fatty acid incorporation into diacylglycerol in HUVEC. Cell Culture. HUVEC were plated in 6 well plates. Cells were grown at 37 °C in a 5% C0₂ incubator with EBM-2 medium. After reaching confluence, medium was changed to Medium-199 + 10% FBS containing 5, 30, and 30 mM glucose+1 mM AICAR. Assessment of mitochondrial membrane potential. After 16 hrs incubation with each medium, the cells were incubated with 50 nM H3-tetraphenylphosphonium (TPP, Amersham, IL) in depolarizing buffer for 15 minutes. The cells were washed with cold PBS 3 times and the H3-TPP counts were measured. Measurement of free fatty acid oxidation and free fatty incorporation into diacylglycerol (DAG) . After 16 hrs incubation, free fatty oxidation was measured by the method described in Example 1. The rate of free fatty acid incorporation into DAG was measured by counting the radioactivities in H3-labeled DAG fraction separated by thin layer chromatography. Incubation of HUVEC with 30 mM glucose for 16 hrs decreased free fatty oxidation by 50% (Figure 8A) and increased free fatty acid incorporation into DAG (esterification) by 40% (Figure 8B) . These changes are accompanied by significant decline in mitochondria membrane potential (Figure 8C) . Treatment with 1 mM AICAR in 30 mM glucose prevented all these changes.

Assessment of the Results

The effects of AICAR on HUVEC in 30 mM glucose were consistent with the scheme depicted in Figure 2, in which a high glucose concentration increases apoptosis by decreasing free fatty-acid oxidation and increasing esterification. AICAR prevented these changes as well as the decreased mitochondrial membrane potential caused by hyperglycemia. Because decreased mitochondrial membrane potential promotes cell apoptosis, this observation further suggests that a decrease in fatty-acid oxidation and an increase in the esterification of fatty- acids cause apoptosis. To test this notion, experiments were performed in HUVEC incubated with 2-bromopalmitate, an inhibitor of CPT1 which inhibits free fatty oxidation, or a higher concentration of free fatty acids (0.5 mM vs. 0.1 mM palmitate). As shown in Figure 9A, inhibition of free fatty acid oxidation by 2- bromopalmitate promotes apoptosis, as does incubation with higher levels of palmitic acid (Figure 9B) . These results are consistent with the proposed metabolic theory of apoptosis and the effects of AICAR to prevent it.

It is understood that those with skill in this and related fields (e.g., synthetic organic chemistry) will identify other agents that increase the oxidation of fatty acids and decrease its esterification (or decrease levels of long chain fatty acyl CoA in the cytosol of the cell) and which might function equally well. Because of the low absorption rate of the AICAR when administered orally

(approximately 5%), compounds which release AICAR after ingestion, methods that will result in the release of AICAR

(e.g., encapsulation), and other methods of administration

(e.g., usage of pumps that deliver small amounts continuously) will need to be explored. In addition, those with skill in the field will identify variations of the invention which are consistent with the disclosure herein.

Claims

CLAIMS We claim:

1. A method for the prevention of vascular disease in mammals associated with metabolic abnormalities, comprising the long-term administration of AICAR.

2. The method of claim 1 wherein said disease is atherosclerotic vascular disease.

The method of claim 1 wherein said mammal is man,

4. The method of claim 1 wherein said disease is selected from the group consisting of blindness, retinopathy and nephropathy caused by diabetic microvascular disease.

5. The method of claim 1 wherein said AICAR is administered via treatment of said mammal with an AICAR precursor.

6. A method for treatment of vascular disease in mammals associated with metabolic abnormalities, comprising the long-term administration of AICAR.

7. The method of claim 6 wherein said disease is atherosclerotic vascular disease.

8. The method of claim 6 wherein said disease is selected from the group consisting of blindness, retinopathy and nephropathy caused by diabetic microvascular disease.

9. The method of claim 6 wherein said mammal is man.

10. The method of claim 6 wherein said AICAR is administered via treatment of said mammal with an AICAR precursor.