Rhonda Oetting Deems

Hypoglycemic effects of a novel fatty acid oxidation inhibitor in rats and monkeys

Increased fatty acid oxidation contributes to hyperglycemia in patients with non-insulin-dependent diabetes mellitus. To improve glucose homeostasis in these patients, we have designed a novel, reversible inhibitor of carnitine palmitoyl-transferase I (CPT I) that potently inhibits fatty acid oxidation. SDZ-CPI-975 significantly lowered glucose levels in normal 18-h-fasted nonhuman primates and rats. In rats, glucose lowering required fatty acid oxidation inhibition of > or = 70%, as measured by beta-hydroxybutyrate levels, the end product of beta-oxidation. In cynomolgus monkeys, comparable glucose lowering was achieved with more modest lowering of beta-hydroxybutyrate levels. SDZ-CPI-975 did not increase glucose utilization by heart muscle, suggesting that CPT I inhibition with SDZ-CPI-975 would not induce cardiac hypertrophy. This was in contrast to the irreversible CPT I inhibitor etomoxir. These results demonstrate that SDZ-CPI-975 effectively inhibited fatty acid oxidation and lowered blood glucose levels in two species. Thus reversible inhibitors of CPT I represent a class of novel hypoglycemic agents that inhibit fatty acid oxidation without inducing cardiac hypertrophy.

Chemosensory function, food preferences and appetite in human liver disease

We evaluated chemosensory function, food preferences, and appetite in 88 patients with liver disease including those with hepatitis, cirrhosis, primary biliary cirrhosis, and sclerosing cholangitis. Reported chemosensory disturbances were common in the patients with liver disease: over 40% reported recent taste changes, and 27% reported recent changes in smell, compared to only 6% of healthy, age- and gender-matched controls with no history of marked chemosensory malfunction. Compared to controls, a greater proportion of liver patients reported food cravings (47 vs. 17%) and food aversions (33 vs. 16%). Foods with a predominantly bitter taste were specifically less preferred in patients with liver disease compared to healthy controls. Patients were also more likely to report poor to fair appetite than controls (37 vs. 5%). Groups of patients with different types of liver pathology showed varying patterns of food preferences, suggesting that dietary recommendations to liver patients might be tailored to the altered preferences associated with a particular type of hepatic dysfunction.

Limonene in expired lung air of patients with liver disease.

As part of an effort to examine the relationship between chemosensory disturbance and oral chemistry, we analyzed expired lung air samples from a series of 24 patients with liver disease and 24 healthy controls using gas chromatography-mass spectrometry. Compared to samples from controls, lung air from patients with liver disease contained unusually high levels of limonene, a monoterpene that is a major component of the essential oil of citrus fruits (0.1 vs 7.0 µg/20 liters for controls and patients). Only half the patients showed high levels of limonene. Patients with noncholestatic liver disease were significantly more likely to have elevated lung air limonene levels than those with cholestatic liver disease (0.2 vs 13.8 µg/20 liters). Responses to food frequency and dietary behavior questionnaires indicated a pattern of diet selection and food preferences that were consistent with a dietary origin for the limonene in these patients.

Pharmacological Strategies for Reduction of Lipid Availability

The chronic elevation of circulating free fatty acid (FFA) levels is associated with insulin resistance.’ In normal individuals, excess energy is converted to FFAs by the liver and stored as triglyceride in adipocytes. The stored energy can be mobilized as needed via an increase in the rate of FFA release from the adipocytes. However, if the balance between storage and release is disturbed such that the FFAs become chronically elevated, insulin resistance is usually manifest. Elevated FFA levels have been implicated in inducing metabolic derangements. For example, over 30 years ago, Randle demonstrated that FFAs have an impact on peripheral glucose utilization in muscle, principally via a FFA-induced suppression of pyruvate dehydrogenase (PDH).* Felber has recently shown that FFAs may be playing a role in glycogen storage3 and Grill has suggested that they have an important role in regulating beta-cell function and thereby altering insulin secretion profile^.^ Furthermore, recent studies by Bergman demonstrate that lowering of circulating FFAs is required for suppression of hepatic glucose production following a meal.5 The causative role of FFAs in the development or maintenance of insulin resistance has been difficult to assess. One reason for this could be that measurement of circulating FFA levels may be an inadequate measure of the role of FFAs since stored triglycerides and local fluxes may be major factors as well. In a normal individual, the stored triglyceride in tissues such as the liver, the muscle, and even the islet is relatively low. However, with chronic elevation of circulating FFA, as occurs in obesity, the triglyceride content of all of these tissues is predicted to be higher than normal, so local storage and release of FFAs may become important. One may ask the following question: “If the circulating FFAs were reduced, would insulin sensitivity improve?” Acute experiments in humans suggest that the answer is no. However, such conclusions may be flawed because the locally stored triglyc-

Relationship between liver biochemical tests and dietary intake in patients with liver disease

Relationships between liver biochemical test values and reported frequency of consumption of various foods were examined using a principal-component analysis of data from 42 patients with chronic liver disease. The statistical procedure identified relationships among biochemical and dietary variables. One relationship included the variables albumin, bilirubin, and frequency of intake of fruits and vegetables, starch, and meats. A relationship was also found between serum alkaline phosphatase (ALP) levels and fat/oil intake. Data from patients with primary biliary cirrhosis (PBC) and noncholestatic liver disease were compared using a correlational analysis. In patients with PBC, serum ALP levels were positively correlated with frequency of intake of fat/oil (r = 0.59, p < 0.01) and meats (r = 0.46, p < 0.05), whereas serum bilirubin (Bili) and aspartate aminotransferase (AST) levels were significantly correlated with frequency of intake of dairy products (rs = 0.48 and 0.45, ps < 0.05 for Bili and AST, respectively), meats (rs = 0.59 and 0.65, ps < 0.01), and fat/oil (r = 0.54, p < 0.02 and r = 0.48, p < 0.05). In patients with noncholestatic liver disease, Bili levels were correlated with frequency of intake of fat/oil (r = 0.58, p < 0.01), and fruits and vegetables (r = 0.68, p < 0.01). These results suggest that the degree of elevation of some liver biochemical tests in patients with liver disease may be affected by dietary intake.

Macronutrient selection in an animal model of cholestatic liver disease.

Diet selection was investigated in an animal model of cholestatic liver disease produced by bile duct ligation. Animals self-selected diets from separate sources of macronutrients (protein, fat, carbohydrate). Diet selection was evaluated when the fat source was comprised of either a primarily medium-chain fat (coconut oil) or a primarily long-chain fat (Crisco vegetable shortening). Relative to intakes of control animals, bile duct ligated (BDL) animals consuming the long-chain fat decreased fat intake, decreased protein intake, and increased carbohydrate intake. Consumption of the fat source was decreased in BDL rats fed the medium-chain fat relative to intakes of control animals, however carbohydrate and protein intakes were not affected. Total caloric intake was comparable to control intakes by day 16 post-ligation in BDL rats fed the long-chain fat and by day 11 in BDL rats fed the medium-chain fat. Body weight gain was significantly greater in BDL rats fed the medium-chain fat than in those fed the long-chain fat. Mortality was 44% in BDL animals fed the long-chain fat, and 0% in those fed the medium-chain fat. The results suggest that BDL animals make dietary selections which may decrease the severity of liver disease. Differences between ligated animals consuming either medium- or long-chain fats suggest that some fat sources may be more beneficial during cholestasis.

Insulin, Sham Feeding, Real Feeding, Glucostasis, and Flavor Preferences in the Rat

In previous work’ we had shown insulin to be capable of reducing the amount of milk that rats would sham feed without producing an aversion to the taste of milk paired with insulin injections. Indeed, in that experiment, animals preferred insulin-paired flavors to saline-paired flavors of milk. In earlier work’ others, however, had found no effect of even large doses of insulin on the real feeding of milk. In this study we returned to these findings and assessed the effects of insulin in realand sham-feeding rats following modest doses of insulin and an overnight fast. In two separate, parallel experiments, rats were allowed to sham or real feed a flavored, sweetened condensed milk solution following an overnight, 17 %-hour fast. Coincident with the offering of the milk were injections of varying doses of insulin (0.4, 08 U/rat) or saline (0 U/rat). Insulin was without effect on milk intake in the real feeding situation, even at doses as large as 5.0 U/rat; however, flavors of milk paired with insulin injections were avoided relative to those paired with saline, at doses greater than 0.8 U/rat. Insulin at these higher dose levels also produced reliable hypoglycemia in these animals (FIG. la for intake and preference data and TABLE 1 for glycemic changes). Insulin significantly reduced the amount of milk sham fed by rats at 0.8 Uhat; however, here the flavors paired with insulin were preferred over those paired with saline (see FIG. lb). Insulin produced a relatively greater hypoglycemia in sham-feeding animals than in those real feeding the milk solutions. Additionally, sham feeding in saline-injected animals did not produce hypoglycemia, consistent with earlier r epor t~ .~ It appears that insulin-paired flavors of milk may be preferred when they are ingested in the absence of a filled stomach or significant gastric emptying (as would occur in the sham-feeding animal when ingested food is draining out of the gastrointestinal system through an open cannula). When real feeding milk, rats seem to avoid flavors paired with insulin injections. Some preliminary work in our laboratory indicates that insulin accelerates gastric emptying, even at the low or modest doses used here. Perhaps, this presents to the small intestine a load too calorically dense and unprepared for absorption, as there is no indication of a repair of the hypoglycemic condition in the real-feeding animal, even at 30 minutes postinsulin (see TABLE 1). If insulin in real-feeding rats is accelerating gastric emptying, an aversive state similar to “dumping”