PLASMA AMMONIA VALUES IN NORMAL AND IN‘ PATHOLOGICAL STATES OF CATTLE, DOGS, AND RATS Thesis for the Degree of M. S.‘ ‘ MICHIGAN STATE UNIVERSITY FRANCES A. KENNEDY 1980 ...... _____ “ECUS L 11:: R A R Y L N ILT‘Iigan saw I, a} University 3 . ~f' —‘ Mu'tvuu" ._. —v OVERDUE FINES: 25¢ per day per item RETURNIIB LIBRARY MATERIALS: Place in book return to remove charge from circulation records ABSTRACT PLASMA AMMONIA VALUES IN NORMAL AND IN PATHOLOGICAL STATES OF CATTLE, DOGS, AND RATS BY Frances A. Kennedy A cation exchange resin technique for plasma ammonia determina- tion was evaluated in experimental and clinical cases of nutritional and hepatic disturbances in cattle, dogs, and rats. This technique's reliability was determined by recoveries and plasma vs serum samples, and by measuring the effect of 5 days of storage. Vitamin E deficient rats fed silver acetate and cod liver oil had a three-fold increase in plasma ammonia in comparison to rats not fed cod liver oil. Rats with higher ammonia values also had hepatic degeneration. After a 6-day starvation cows fed a high-protein ration had higher plasma ammonia values than cows fed a low-protein ration. Plasma ammonia values in 4 dogs and 1 cat with hepatic disturbance were higher than control subjects. Plasma ammonia values were lower 30 minutes after feeding calves a diet containing up to 1% urea. Elevated plasma ammonia as determined by the cation exchange technique was a reliable indicator of hepatic injury in animals. PLASMA AMMONIA VALUES IN NORMAL AND IN PATHOLOGICAL STATES OF CATTLE, DOGS, AND RATS BY Frances A. Kennedy A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Pathology 1980 LJ'I‘1‘7 7/ ACKNOWLEDGEMENTS I wish to thank Dr. C. K. Whitehair for his constant assistance as chairman of my guidance committee. I also wish to thank Dr. R. F. Langham and Dr. G. L. Waxler of the Department of Pathology and Dr. J. B. Dalley of the Department of Small Animal Surgery and Medicine for their guidance and advice during this research. I am grateful to Dr. A. Telles, Dr. C. W. Lopes, Dr. R. B. DaSilva, Dr. T. P. Mullaney, and Dr. J. Spalding for their help in obtaining specimens for this research. I am sincerely grateful to Dr. E. Roege for the generous offer of her time in preparing electron microscopic data. The assistance of research technicians Linda Stegherr and Melissa Blue and research animal caretaker John Allen is also greatly appreciated. ii TABLE OF CONTENTS Page INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . 1 LITERATURE REVIEW. . . . . . . . . . . . . . . . . . . . . . . . 5 Biochemistry of Ammonia Metabolism. . . . . . . . . . . . 5 Hepatic Encephalopathy. . . . . . . . . . . . . . . . . . ll Reye's Syndrome . . . . . . . . . . . . . . . . . . . . . l7 Portosystemic Anastomosis . . . . . . . . . . . . . . . . l9 Ammonia Metabolism in Malnutrition. . . . . . . . . . . . 23 Ruminant Ammonia Metabolism . . . . . . . . . . . . . . . 29 Methods of Blood Ammonia Determination. . . . . . . . . . 35 Summary . . . . . . . . . . . . . . . . . . . . . . . . . 38 OBJECTIVES . . . . . . . . . . . . . . . . . . . . . . . . . . . 40 MATERIALS AND METHODS. . . . . . . . . . . . . . . . . . . . . . 41 Blood Ammonia Determination . . . . . . . . . . . . . . . 41 Recoveries and Serum vs Plasma. . . . . . . . . . . . . . 45 Vitamin E Deficiency in Rats. . . . . . . . . . . . . . . 46 Bovine Starvation . . . . . . . . . . . . . . . . . . . . 47 Dietary Urea. . . . . . . . . . . . . . . . . . . . . . . 47 Clinical Cases. . . . . . . . . . . . . . . . . . . . . . 48 Histopathologic Technique . . . . . . . . . . . . . . . . 48 RESULTS 0 O O O O O O O O O O O O O O O O O O O O O O O O O O C . 50 Blood Ammonia Determination Technique . . . . . . . . . . 50 Vitamin E Deficient Rats. . . . . . . . . . . . . . . . . 50 Bovine Starvation . . . . . . . . . . . . . . . . . . . . 54 Dietary Urea. . . . . . . . . . . . . . . . . . . . . . . 59 Clinical Cases. . . . . . . . . . . . . . . . . . . . . . 61 Case 1 . . . . . . . . . . . . . . . . . . . . . . 61 Case 2 . . . . . . . . . . . . . . . . . . . . . . 65 Case 3 . . . . . . . . . . . . . . . . . . . . . . 65 Case 4 . . . . . . . . . . . . . . . . . . . . . . 66 Case 5 . . . . . . . . . . . . . . . . . . . . . . 66 DISCUSSION . . . . . . . . . . . . . . . . . . . . . . . . . . . 68 Blood Ammonia Determination Technique . . . . . . . . . . 68 Vitamin E Deficient Rats. . . . . . . . . . . . . . . . . 69 Bovine Starvation . . . . . . . . . . . . . . . . . . . . 70 Dietary Urea. . . . . . . . . . . . . . . . . . . . . . . 71 Clinical Cases. . . . . . . . . . . . . . . . . . . . . . 72 Page SUMMARY. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74 BIBLIOGRAPHY C O O O O O O O O O O O O O O O O O O O O O O O O O 75 iv Table LIST OF TABLES Recoveries of ammonia standards added to human plasma 0 O O O 0 O O O C O O O O C O O C O O O C O O O O Ammonia levels in simultaneously run human plasma and serum samples . . . . . . . . . . . . . . . . . . . . . Changes in ammonia levels in plasma samples held at -20 C O O O O O O O O O O O O O O O O O O O O O I O O 0 Plasma ammonia levels in Sprague-Dawley rats with experimentally induced liver damage . . . . . . . . . . Effect of 6 days of starvation on bovine plasma ammonia levels. 0 O O O O O O O O O O O O O O O O O O O O O O 0 Plasma ammonia levels (Hg/100 ml) of calves fed increasing amounts of urea. . . . . . . . . . . . . . . Blood urea nitrogen levels (mg/100 ml) of calves given dietary urea. . . . . . . . . . . . . . . . . . . Laboratory results of patients suspected of having hepatic disease . . . . . . . . . . . . . . . . . . . . Blood ammonia concentrations (Hg/100 ml) of patients suspected of having hepatic disease . . . . . . . . . . Page 51 52 53 55 58 6O 62 63 64 LIST OF FIGURES Figure Page 1 Biochemical pathways of ammonia metabolism. . . . . . . . 8 2 Possible sites of ammonia toxicity in the brain . . . . . l6 3 Effect of ammonia on malate—aspartate shuttle in the brain . . . . . . . . . . . . . . . . . . . . . . . . 24 4 Light micrographs of liver of a rat given a vitamin E deficient diet, silver acetate, and cod liver oil . . . . 57 5 Electron micrographs of liver of a rat given a vitamin E deficient diet, silver acetate, and cod liver oil . . . . 57 vi INTRODUCTION Ammonia (NH3) has long been suspected of having a key role in the development of central nervous system (CNS) disturbance in rela- tion to metabolic disease. The liver acts as a detoxification center for many of the body's metabolites, including NH It is during 3. episodes of liver failure that patients experience signs referable to the brain, giving rise to the term hepatic encephalopathy (HE). A fairly constant laboratory finding in these patients is elevated blood NH3 levels (Bessman and Bessman, 1955; Schenker et al., 1974; Snyder, 1978; Walker and Schenker, 1970). Experimental confirmation of ammonia's part in pathogenesis of HE has been thwarted by the unreliability of clinical laboratory determinations of blood NH3 values. Much work, therefore, has been aimed at developing an accurate technique for blood NH3 determination vfifixfli would be con- venient for use in a clinical laboratory. Most recently, attention has turned to the possible role of NH3 in the pathogenesis of Reye's syndrome. As early as the 1920's, a syndrome of fatty degeneration of the liver accompanied by severe neurological disturbance was reported (Brain et al., 1929). It was not until much later that the disease and its possible pathogenesis were more fully characterized (Reye et al., 1963). The relationship between the brain and liver having been recognized, the body's 2 metabolites (most specifically NH3) have been considered the possible cause of acute brain swelling seen in Reye's syndrome (Huttenlocher et al., 1969). Experimental data are being accumulated to support this hypothesis (Altenau and Kindt, 1977; Kindt et al., 1977). While Reye's syndrome is associated with acute liver failure, portocaval shunts (PCS) are responsible for a chronic deviation in the body's normal detoxification system. Portocaval shunts are being reported with increasing frequency in veterinary literature (Beech et a1, 1977; Cornelius et al., 1975; Prouty, 1975; Rogers et al., 1977; Vitums, 1961; Vulgamott, 1979). Through dietary control or surgical correction (Strombeck et al., 1977), these patients are some- times maintained so that permanent CNS damage due to elevated intra- cranial NH3 does not occur. On the other hand, a PCS is sometimes created surgically in an effort to control ascites (Keefe et al., 1961). Following such surgery the clinician must deal with the same problems seen in a naturally occurring PCS. Experimentally produced PCS have been created to evaluate the metabolic alterations in liver and brain resulting from altered liver blood flow (Hindfelt et al., 1977; Silen et al., 1957; Starzl et al., 1973). Consideration of ammonotelic metabolism should not be limited to the brain, however. Ruminants can use non-protein nitrogen (NPN) sources in lieu of more expensive dietary protein. It has been shown that microbial use of NH from urea, for protein synthesis in the 3, rumen can be an efficient method of providing the ruminant with essential amino acids (Hogan and Weston, 1970; Loosli, 1949; Smith, 1975). Optimum levels of NPN have been sought to achieve maximum bacterial protein metabolism (Satter and Slyter, 1974; Slyter et al., 1973). With the overzealous use of NPN in the ruminant diet, however, urea toxicity can result. Hydrolysis of urea to NH3 has been shown to proceed more rapidly than microbial utilization of NH for protein 3 synthesis (Bloomfield et al., 1960). The consequence of excess urea intake, therefore, is an elevation of rumen NH3 (Ciszuk, 1973; Hillis et al., 1971) and subsequent hyperammonemia (Hogan, 1961). Treat- ment of urea toxicity has been aimed at prevention of hyperammonemia, with some studies being done on the possible effects on bovine repro- ductive performance (Word et al., 1969). Some research has also been carried out on the possible effect of NH3 toxicosis during rehabilitation of malnourished individuals (Stevens et al., 1975). Fairly distinct neurological signs have been produced in dogs given protein deficient diets (Stewart and Platt, 1968). Neurological signs have been seen in children recovering from kwashiorkor. Some clinicians have suspected NH3 as a contributor to these signs. Again, however, the difficulty and unreliability of blood NH3 determinations have limited studies of these cases (Balmer et al., 1968). A balanced amino acid replacement has been found to be very important in recovery of these patients. An arginine- deficient diet given either therapeutically (Heird et al., 1972) or experimentally (Morris and Rogers, 1978) has resulted in CNS disturbance associated with hyperammonemia. The central problem in all these areas of study has been a lack of a trustworthy method of measuring blood NH Microdiffusion tech- 3. niques were introduced in the 19305 (Conway, 1935) but were found to be cumbersome and variable in their results (Acland and Strong, 1968). Techniques in which blood is deproteinized have also been in use clinically (McCullough, 1967). This technique requires whole blood, however, so the values obtained are uniformly high. An enzymatic 4 method of blood NH3 determination has recently been developed (DaFonseca-Wollheim, 1973). This technique is quite expensive, making it of limited value as an experimental technique. A cation exchange resin technique has shown the most promise as a reliable means for determining plasma NH on large numbers of blood samples (Dienst, 1961; 3 Hutchinson and Labby, 1962; Miller and Rice, 1963). Since no research in the area of NH3 metabolism could be carried out without a serviceable blood NH determination technique capable 3 of handling large numbers of samples, study into the area of labora- tory technique was highly warranted. It was for this reason that work was begun using blood samples from various clinical and experimental cases of liver damage and dietary alteration. The cation exchange resin technique was used for plasma NH3 determination. LITERATURE REVIEW Much recent literature has been devoted to metabolism of NH3 and its effect on the pathogenesis of disease. This review outlines those articles most pertinent to this research, with particular attention paid to NH metabolism in the CNS, ruminant utilization of NH , and 3 3 recent developments in blood NH determination techniques. 3 Biochemistry of Ammonia Metabolism Ammonia is an essential nutrient necessary for the biosynthesis of protein. As important as it is in normal metabolism, its excess can cause overloading of the processes of protein and nucleic acid synthesis, resulting in accumulation of free NH (Committee, 1979). 3 Ammonia belongs to the class of weak electrolytes which exist in part as undissociated molecules. The degree of ionization of NH governs 3 its behavior in chemical reactions, adsorption to surfaces, and pene- tration of membranes (Visek, 1968). Mammals take in dietary protein in excess of that needed for protein synthesis. Americans, for example, use amino acids for 10 to 25% of caloric needs. During amino acid degradation, NH is released in the intestinal lumen (Committee, 1979). 3 The NH3 is absorbed and travels via the portal circulation to the liver, where it is removed from the blood. Even when 70% of the liver has been removed, mammals can tolerate higher than normal blood NH3 concentrations due to the efficiency of the liver's NH3 detoxification mechanism (Visek, 1968). Most catabolism of absorbed amino acids takes 6 place in the liver (Lehninger, 1970), in which various biochemical pathways work to remove free NH from the tissue (Figure 1). 3 While the liver is working to remove NH from the blood, the 3 kidney is usually a net NH producer. The major source of NH pro- 3 3 duced in the kidney is thought to be glutamine (Lewis, 1976). In acidosis, glutaminases in renal epithelium release NH from glutamine. 3 The NH3 molecule then accepts hydrogen ions to form ammonium ion + . (NH4), which is excreted in the urine. Alkalosis associated Wlth hypokalemia (often induced by diuretics) increases net NH3 formation (Committee, 1979). In hypokalemia, extracellular alkalosis combined with intracellular acidosis favors entry of NH3 into cells. This augments the pathogenesis of hepatic coma seen in hyperammonemia (Lewis, 1976). Normal intracellular pH in skeletal muscle is 7.0 compared with 7.4 for extracellular pH. This pH difference favors + diffusion of NH3 into the cell so that intracellular NH3-NH4 levels are 2.5 times those seen extracellularly (Visek, 1968). This diffusion ~11 rate has been calculated, using red blood cells (RBC), to be 3.6 x 10 moles NH3/sec for each RBC. Once inside the cell, the NH3 is instantly + (Klocke et al., 1968). ionized to the poorly diffusible ion NH4 In the liver, a-ketoglutarate is the ultimate acceptor of amino groups from most of the other amino acids. It can also accept NH3 in the following reaction: glutamic acid _ + dehydrogenase a-ketoglutarate + NAD(P)H + H + NH glutamic acid 4 + + NAD(P) + H20 The reaction is pulled to the right, so that NH is trapped. Because 3 of this, normally only small amounts of NH can exist in the presence 3 of a-ketoglutarate. This reaction is reversible, however, so that NH3 7 can either be bound or released. Ammonia is more effectively trapped in the rapid reaction: glutamine . . s h . . Glutamic aCid + NH3 + ATP ynt etagg glutamine + ADP + P1 The reaction equilibrium lies well to the right, resulting in low levels of free NH3. The glutamine molecule has no net charge and so can move freely through membranes. It is the only molecule other than glucose that can freely cross the blood-brain barrier in substantial quantities. Once inside a cell, glutamine can release glutamic acid and NH3. These molecules are less permeable because at physiological pH only + 1% of total NH3-NH4 exists as the uncharged, diffusible NH3 molecule (Committee, 1979). While glutamine synthesis is the major fate of exogenous NH3, If NH urea formation is responsible for the final excretion of NH3. 3 is given intravenously to a dog, urea can be formed at a rate of 2 mg of nitrogen per kilogram per minute (Committee, 1979). Elevated blood NH levels will normally stimulate urea synthesis, but excessive NH 3 3 can inhibit respiration and substrate utilization in the liver (Katunuma et al., 1966). Cases of urea cycle enzyme deficiency have been reported. Congenital arginosuccinate synthetase deficiency in a dog resulted in hyperammonemia because of inability of the liver to convert NH3 from the intestines to urea for excretion (Strombeck et al., 1975). Of particular interest is metabolism of NH3 in the brain. Not all enzymes necessary for the various steps of NH3 detoxification are present in the brain (Figure 1). This, combined with the fact that hyperammonemia is most dramatically evidenced in CNS signs, makes .Emflaonmug mazes—Em mo menmzfimm Hmogmnooflm .H musmwm .mzo 2: 5 Bow 222.02. 3 :32? 2:32 on >2: .30:an no.2: 05:53.0 l I l I 553.23 moEaucu @ _ // 05:35.0 E+mo< 28 226 _ 2. 5.50 a 0.53:» E _ _ / wanna—”.3” 10:35 _ / / nh< _ ® . a a 05:35 _ Bow 2:. u .0 / / 2233.3 _ + oEu€uom:E.o£5.Eo Q / _ 380553 22.32:. 3:3an 9 / n m T— Z 9 discussion of NH3 biochemistry in the brain most important. Brain NH3 levels have been shown to be elevated in excitatory states, such as convulsions. On the other hand, depressed states of brain neuronal activity, such as anesthesia or hibernation, have lower brain NH3 con- centrations. It has also been shown that brains taken from immature rats (less than 3 weeks old) have twice the tissue NH3 concentration as seen in mature rat brain (Tsukada, 1972). Elevated tissue NH3 levels, however, may be interpreted as either cause or effect of altered brain function in these conditions. More precisely, clinical signs of hepatic encephalopathy have been thought to be related to elevation of blood and brain NH3 levels. These signs, including asterixis (flapping tremor), decerebrate rigidity, hyperpnea, and coma, have clinically been related to dysfunction in the base of the brain. Because of this, investigation has been carried out to analyze the dynamics of the brain's handling of excess NH3 levels and to locate specific areas of the brain most altered by toxic effects of NH3 (Schenker et al., 1967). Glutamine synthetase activity is localized in the extramito- chondrial cytoplasm of glial cells of the brain (Committee, 1979). In hyperammonemic states the synthesis of glutamine has priority over other energy-consuming reactions that bind free NH3 in the brain (Tsukada, 1972). Glutamic acid dehydrogenase levels are elevated in brain tissue from experimental cases of hyperammonemia. As would be expected, elevated blood NH reduced brain glutaminase levels, and so 3 suppressed the release of free NH from glutamine. When hyperammonemia 3 was combined with subacute liver injury due to carbon tetrachloride, however, glutamic acid dehydrogenase levels decreased and glutaminase levels increased in brain. It would appear, therefore, that normal 10 liver function is necessary to maintain these enzymes at proper levels in the brain for NH3 detoxification (Kisfaludy, 1975). The molecular kinetics of the hyperammonemic state have been studied using brain slices in varying tissue media. When brain slices are incubated in a medium with high levels of potassium, NH3 + + accumulation decreases. This suggests that NH4 and K compete for the same ion transport across brain cell membranes. Ammonia formation is also depressed by high levels of glucose in the medium. This has been interpreted as due to either less proteolysis with more available glucose or an NH binding mechanism (Tsukada, 1972). The effect of 3 excess NH3 on brain oxygen consumption has had varying results. Oxygen consumption has been reduced in brain slices by high levels of NH3 in some laboratories (Tsukada, 1972; McKhann and Tower, 1961), while another laboratory found no alteration in brain 02 consumption with high NH concentrations in the tissue media (Schenker et al., 3 1967). Alteration of intracellular pH has been blamed for the changes in biochemical reactions in brain tissue. It has been shown, however, that if carbon dioxide tensions are kept relatively stable in brain tissue, intracellular acidosis seen in NH3 intoxication is offset by transport of basic NH molecules into the cells (Hindfelt and 3 SiesjbI, 1971). As with O2 consumption, adenosine triphosphate (ATP) levels in brain tissue with elevated NH3 concentrations have been found either to be unchanged (Hindfelt and SiesjBII, 1971) or decreased (Schenker et al., 1967). These results still leave the possibility of a small, isolated, critical area of ATP depletion which may be immeasurable due to its relation to the rest of the brain chemistry. ll Hepatic Encephalopathy» The most devastating effects of deranged NH3 metabolism are generally seen in CNS disturbances. While there is still controversy as to its pathogenesis, hepatic encephalopathy (HE) is believed by most workers to be caused by abnormal NH3 metabolism. One researcher has even found a case of post-cirrhotic encephalopathy in Shakespearean literature. Sir Andrew Aquecheek (Twelfth Night; Or What You Will, 1602) was noted for his alcoholism and deranged intellect and said: Methinks sometimes I have no more wit than a Christian or an ordinary man has; but I am a great eater of beef, and I believe that does harm to my wit. Shakespeare seems to have unknowingly constructed a case of hepatic cirrhosis complicated by protein intoxication (Summerskill, 1955). It is known today that digestion of meat releases NH3, which is absorbed and carried via the portal vein to the liver. In cirrhosis, not only is the liver unable to metabolize NH to form urea, but 3 portal vessels shunt around the fibrotic liver and carry high levels of NH3 into the blood (Snyder, 1978). While abnormal liver function is a prerequisite for the develop- ment of HE, certain conditions can precipitate coma in individuals with compensated liver function. Any protein digestion in the intes- tinal tract results in NH3 release. While ingested meat protein is the most common source, gastrointestinal hemorrhage and constipation also increase amount and duration of protein availability in the intestinal lumen (Snyder, 1978). Kidney failure may also contribute when blood urea levels become elevated. Urea can freely diffuse from the blood into the intestinal lumen. There urea is hydrolyzed by bacteria to release NH3 for absorption. This NH3 must then be cycled back to the liver for urea formation (Committee, 1979). If liver 12 function is impaired, blood NH levels become markedly elevated. 3 Inborn defects in the urea cycle may also result in intermittent CNS signs reSembling HE. These defects include congenital arginosuccinic aciduria, citrullinuria, arginase deficiency, congenital lysine intolerance, and rare cases of carbamyl phosphate synthetase or ornithine transcarbamylase deficiency (Lewis, 1976). Hepatic encephalopathy can present with a variety of clinical manifestations. In humans, dementia (mental deterioration) is the earliest sign, with the patient showing increased irritability, untidiness, apathy or altered sleep rhythm (Schenker et al., 1974). These subtle signs will progress to rigidity and tremor. Rare focal CNS changes, including irreversible paraplegia, have been reported (Lewis, 1976). The most characteristic clinical sign of HE is asterixis, or "flapping tremor." Here the fingers are held laterally with flexion and extension of the metacarpophalangeal and wrist joints every one to two seconds. This is usually bilateral but asynchronous (Schenker et al., 1974). If this progresses to decere- brate rigidity and coma, the prognosis is very poor. The patient's condition may improve, but usually the CNS signs progress. Hepatic encephalopathy has been categorized by the Modified Parsons-Smith Criteria as follows: Grade 0: no abnormality detected. Grade 1: subtle personality change, trivial lack of aware- ness, shortened attention span, impairment of addition and subtraction. Grade 2: lethargy, facade if personality present, but con- fused, obvious personality change. Grade 3: very confused, semistupor or somnolent, gross disorientation. Grade 4: coma (Snyder, 1978) 13 In some respects HE clinically resembles the encephalopathy seen in Wilson's disease. However, the latter is characterized by familial occurrence with abnormalities in copper metabolism (Victor et al., 1965). Also, the fact that HE patients do sometimes recover suggests a metabolic rather than a structural defect (Schenker et al., 1974). Clinical laboratory data are variable in cases of HE. An ele- vated blood NH value is often seen, while an exaggerated blood NH 3 3 elevation in response to oral NH challenge is more specific (Strombeck 3 and Gribble, 1978). Some workers feel there is a direct relationship between severity of CNS signs and blood NH levels (Victor et al., 3 1965; Bessman and Bessman, 1955), while others feel the correlation is not so reliable (Schenker et al., 1974; Eiseman et al., 1955). Electroencephalographic changes may not be diagnostic and some liver function tests may be normal. Hypoglycemia is often seen, probably due to altered glycogen metabolism in the liver. Elevated cerebro- spinal fluid (CSF) glutamine is a fairly consistent finding (Schenker et al., 1974). Histopathologic liver lesions seen in HE will depend on the particular case, since many hepatic alterations can produce HE. Encephalopathy has been reported in connection with cirrhosis (Snyder, 1978), chronic active hepatitis (Strombeck and Gribble, 1978), and obstructive jaundice, where impairment of bile excretion of NH3 was considered to be contributory to hyperammonemia (Meyer et al., 1980). A separate section of this review will cover anomalies of portal circu- lation which also cause hyperammonemia and HE. Pathological changes in the brain vary. One report outlines cases in which "bilateral foci of degeneration and astrogliosis" in the basal ganglia were the most prominent changes (Victor et al., 1965). Others report l4 degeneration of the white matter in HE (Hoerlein, 1971). Changes in protoplasmic astrocytes are by far the most consistently reported alteration in both human and veterinary medical literature (Schenker et al., 1974; Lewis, 1976; Strombeck et al., 1975; Hoerlein, 1971; Sherding, 1979; Beech et al., 1977). This altered cell has been designated the Alzheimer type II astrocyte. By light microscopy the cell has a pale, vacuolated, twisted nucleus with little cytoplasm. The astrocyte is thought to regulate brain extracellular fluid K+ concentrations. It has been theorized that, since the K+ and NH: ions have the same charge as well as approximately the same radius, . + . . + . elevated brain NH may interfere Wlth K flux into and out of the 4 astrocyte (Lewis, 1976). By far the greatest controversy has centered on the possible pathogenesis of HE. Since the liver is capable of synthesizing cerebral stimulants as well as detoxifying cerebral depressants (Committee, 1979), the lack of the liver's normal function would ultimately result in altered brain function. So—called "liver tropic factors" are thought to be necessary for normal brain metabolism, and their lack results in the brain being more susceptible to injury (Lewis, 1976). If a brain is maintained in a brain perfusion system with no liver in the system, normal cerebral metabolic and electrical activity will cease. If, however, the nucleotides cytidine and uridine, which are normally liberated from the liver, are added, normal cerebral funCtion is maintained (Walker and Schenker, 1970). The most durable hypothesis for the pathogenesis of HE seems to be the energy depletion theory (Bessman and Bessman, 1955). This theory proposes that a-ketoglutarate is depleted in the formation of glutamine in NH3 detoxification. Alpha-ketoglutarate is needed in the brain's 15 Krebs cycle for energy production. Its use in NH3 detoxification, therefore, blocks cerebral energy synthesis (Figure 2). This hypothe- sis is supported by the finding of elevated brain glutamine and low brain stem ATP and phosphocreatine in HE (Lewis, 1976). While the mechanism is controversial, most workers believe impaired energy metabolism at least contributes to some degree to the development of HE (Schenker et al., 1974; Walker and Schenker, 1970; Eisman et al., 1955). Ammonia is not the only compound implicated in the pathogenesis of HE. Short chain fatty acids (butyric, valeric, octanoic) released by digestion of dietary fats are thought to have a direct toxic effect on neuronal synaptic membranes. Biogenic amines (octopamine, B-phenylethanolamine) from intestinal amino acid degradation are thought to act as weak neurotransmitters, blocking normal synaptic transmission (Sherding, 1979). Disturbed amino acid metabolism is also thought to play a role in HE. Normally, plasma branched-chain amino acids (valine, leucine, isoleucine) are at approximately three times the level of aromatic amino acids (phenylalanine and tyrosine). Insulin tends to keep plasma branched-chain amino acids at low levels, with glucagon having the opposite effect. While insulin's influence tends to dominate, the liver metabolizes aromatic amino acids to keep the ratio at 3:1 (branched-chain:aromatic). When liver function is disturbed, however, plasma aromatic amino acid levels rise. Because branched-chain and aromatic amino acids compete for the same transport mechanism into the brain, this allows more aromatic amino acids than normal into the brain. Aromatic amino acids are precursors for the inhibitory neurotransmitter serotonin, while branched-chain amino acids are necessary for the formation of excitatory neurotransmitter 16 GIucose-——) _Glucose-6-P Choflne acetylase Phosphofructoklnase ATP Acetylchollne NAD NADH °"°“"° PYWVBAB ———) Acety+l CoA-—-7 L—Acetoacetate Lactate Oxaloacetate Citrate Succlnate "AD” LATP FORMED] MAD; Succinyl CoA t9 a-ketoglutarate y-Aminobutyrate ‘ NADH ® Glutamine NAD synthetase NH J Glutamic acid \L AT: ) Glutamine 002 G) ® Oxidative decarboxylation ot pyruvate impaired % NADH depletion a -ketoglutarate depletion ATP used to form glutamine Figure 2. Possible sites of ammonia toxicity in the brain. l7 norepinephrine. The overall effect, therefore, is CNS depression (Strombeck and Rogers, 1978; Sherding, 1979). Reye's Syndrome Of recent interest is the occurrence of an acute CNS disturbance following flu-like symptoms. This disorder was designated "encephalo- pathy and fatty degeneration of the viscera" by Reye and his associates in 1963 but was later called simply Reye's syndrome. In most cases patients seemed to be recovering from an initial illness resembling a cold or flu, when severe vomiting began accompanied by high fever (Huttenlocher et al., 1969). The patient first was very excited, exhibiting hyperpnea and tetanic spasms followed by convulsions in severe cases. A characteristic posture was seen in these patients, with elbows flexed, legs extended, and hands clenched. Average survival time was 27 hours from the onset of symptoms. Those that survived, however, returned to normal function with no noticeable impairment of CNS function (Reye et al., 1963). After Reye's report, cases were reported from other hospitals in which an identical syndrome had been seen and attributed to ingested toxins (Elliot et al., 1963; Brain et al., 1963). Attention was then brought to a previous report in which an epidemic of six cases of the syndrome were described as "acute meningo-encephalitis of childhood", with patients ranging in age from 3 to 18 years (Brain et al., 1929). Clinical laboratory data on these patients included low glucose levels in both blood and CSF with elevated serum glutamic pyruvic transaminase (SGPT) and serum glutamic oxaloacetic transaminase (SGOT). A somewhat delayed prothrombin time has also been reported (Reye et al., 1963; Huttenlocker et al., 1969). An aid in differentiating this from other liver disorders is a normal serum bilirubin level. The 18 most striking and consistent clinical pathological abnormality, how- ever, is an elevation in blood NH3 level (Huttenlocker et al., 1969). There was a report of one case in which lumbar puncture to obtain CSF brought about a remarkable remission of signs, suggesting the possi- bility of elevated CSF pressure in these cases (Brain et al., 1929). Autopsies of patients having these signs have revealed cerebral swelling grossly. Microscopically, neurons in the cerebral cortex have undergone eosinophilic necrosis, sometimes in the laminar pattern seen after convulsions of any cause (Reye et al., 1963; Huttenlocker et al., 1969; Brain et al., 1929). Edema of cerebral tissue has also been reported (Brain et al., 1929; Huttenlocker et al., 1969). In the areas of neuron necrosis, astrocyte nuclei are sometimes swollen (Huttenlocker et al., 1969). but inflammatory infiltrate is not seen (Brain et al., 1929; Reye et al., 1963; Huttenlocker et al., 1969). The liver in cases of Reye's syndrome is grossly enlarged, firm, and bright yellow and the renal cortex is slightly widened and pale (Reye et al., 1963). Microscopically, there is diffuse fatty change in the liver and the proximal convoluted tubules of the kidney (Huttenlocker et al., 1969). Some cases also have fatty change of the myocardium and pancreatic acinar cells (Reye et al., 1963). The liver does not have evidence of inflammatory infiltrate and resembles the liver of a child dying of starvation or of any disease which results in virtual starvation. It does not resemble the liver seen in acute, fulminant hepatitis, in which case the liver is usually shrunken with pronounced necrosis and inflammatory infiltrate (Huttenlocker et al., 1969). In sorting out the clinical signs and lesions seen in Reye's syndrome, certain insight can be gained into its possible pathogenesis. 19 The vomiting seen at the onset may be the result of excessive CNS stimulation. Hyperpnea was initially thought to be a response to metabolic acidosis. This is accompanied by respiratory alkalosis, however, which would indicate overcompensation for any acidosis in the system. It is therefore believed that hyperpnea is due to excessive stimulation of respiratory centers in the medulla (Huttenlocker et al., 1969). Because of the consistent elevation of blood NH3 levels, researchers have tried to reproduce Reye's syndrome by injecting intravenous ammonium acetate into experimental animals. Cats were given NH along with Evans blue dye in order to visualize any extrava- 3 sation of fluid from cerebral vasculature. In these experiments edema was not seen, as defined by increased tissue water (Kindt et al., 1977). When Rhesus monkeys were given intravenous NH cerebral 3, blood flow increased markedly to the point that autoregulation and response to CO2 stimulation were lost. Grossly, cerebral vessels were engorged during the experiment. These researchers hypothesized that reduced cerebral energy metabolism and tissue lactacidosis would stimulate vasodilatation. Another possibility would be a direct toxic effect of NH3 on vessel walls, inhibiting their normal response (Altenau and Kindt, 1977). Portosystemic Anastomosis Veterinary literature has devoted a great deal of discussion to the subject of anomalous anastomoses between the portal vein and the systemic circulation. ,Normally only about 30% of the liver's blood supply arrives via the hepatic artery, with the portal vein as the source of the remaining 70%. Along with its purpose of supplying nutrients to the liver, the portal vein carries the by-products of bacterial and enzymatic degradation of ingested material from the 20 intestine to the liver for detoxification. The colon in particular produces large quantities of NH3 which must be converted to urea in the liver. When the portal vein bypasses the liver and diverts its flow into the systemic circulation, blood NH levels become markedly 3 elevated, especially after a meat meal. This NH3 then may lead to encephalopathy not unlike that seen in liver failure, with stupor, incoordination, behavioral changes and convulsions the most often reported signs (Cornelius et al., 1975). Clinicians consistently find an unusually small liver in these animals (Sherding, 1979; Vulgamott, 1979; Cornelius et al., 1975; Prouty, 1975) when the shunt is the primary lesion. Ascites is also recognized in some cases when portal hypertension or hypoproteinemia results (Rogers et al., 1977). Fasting blood NH is usually elevated in animals with portocaval 3 shunts (Vulgamott, 1979; Beech et al., 1977; Cornelius et al., 1975). An NH3 tolerance test should be carried out, however, for a definitive diagnosis. Thirty minutes after oral administration of 100 mg/kg NH Cl, blood NH 4 3 values will rise 300 to 400% above fasting levels in animals with portocaval shunts (Sherding, 1979). Other clinical pathological data will vary depending on the extent of liver damage resulting from the abnormal blood supply to the liver. Slight eleva- tions in SGPT and sulfbromphthalein retention have been seen, apparently due to chronic impairment of hepatocyte metabolism and reduced blood flow. Ammonium biurate crystals are also frequently found in the urine (Cornelius et al., 1975). Anatomical alterations in portal circulation can be either acquired or congenital. Acquired portosystemic shunts are usually the result of obstruction of the portal vein. Accounts in veterinary 21 literature have described such obstruction due to peritoneopericardial herniation of the liver (Vulgamott, 1979), cirrhosis, malignant lymphoma (Vitums, 1961), and portal vein thrombosis (Beech et al., 1977). These obstructions result in portal hypertension and dilata- tion of normally nonfunctional communications between portal and systemic veins, known as varices. In dogs and cats these communica- tions have been identified as the following: 1. gastrophrenic collaterals (gastric veins to phrenico- abdominal veins) 2. pancreaticoduodenal 3. splenorenal collaterals (splenic vein to left renal or gonadal veins) 4. mesenteric collaterals (mesenteric vein to left renal or gonadal vein) 5. hemorrhoidal collaterals in the sacral region 6. omental collaterals involving paraumbilical or ventral abdominal veins [often secondary to omental adhesions] (Vulgamott, 1979) Care must be taken in identifying one of these shunts as acquired, however, since congenital dilatation of these venous varices sometimes occurs (Thrall, 1980). Portocaval shunts have also been created surgically for treatment of chronic ascites due to heart failure in dogs (Keefe et al., 1961). More common than these acquired shunts are those present at birth due to a congenital abnormality (Prouty, 1975; Rogers et al., 1977). In animals these anomalies include the following: 1. persistent patency of fetal ductus venosus 2. atresia of portal vein with functional collateral portosystemic shunts 3. anomalous connection of portal vein to the caudal vena cava caudal to liver 4. anomalous connection of portal vein to the azygous vein 5. drainage of the portal vein and caudal vena cava into the azygous vein (Cornelius et al., 1975) 22 Histological changes in these livers are much less revealing than the gross lesions. Generally, centrolobular congestion, degeneration, and atrophy are found (Vulgamott, 1979) along with periportal fatty change and fibrosis (Beech et al., 1977; Cornelius et al., 1975). In a horse with an acquired portocaval shunt, hyperammonemia, and encephalopathy, Alzheimer type II astrocytes were found in the basal ganglia and cerebral cortex (Beech et al., 1977). While treatment would seem a simple matter of ligating the shunt- ing vessel, this results in acute portal hypertension and death (Sherding, 1979). Instead, a method of reducing the shunt lumen size by approximately 80% has given good results in a dog with shunting between mesenteric vessels and the caudal vena cava (Strombeck et al., 1977). If surgery is impossible, treatment is essentially the same as used in most patients where chronic hyperammonemia is a problem. This includes a low protein diet and neomycin orally to reduce intestinal bacterial flora. A synthetic disaccharide, lactulose, has also proven effective in lowering intraluminal pH. This keeps the ammonia molecule in the ionized state which is not readily absorbed (Cornelius et al., 1975). As stated earlier, livers in animals with portosystemic anasto- mosis are smaller than normal. In vascular transposition experiments it was found that this is not due entirely to a reduced total blood volume into the liver. Instead, it was discovered that pancreatic hormones, especially insulin, could stimulate hepatocyte hypertrophy and hyperplasia. If, however, pancreatic venous flow is diverted from a particular liver lobe, this area will atrophy (Starzl et al., 1973; Silen et al., 1957). The effect on brain metabolism in portocaval shunts has been studied in an effort to further clarify the mechanism 23 of hepatic encephalopathy. Rats with surgically created portocaval shunts are much more susceptible to elevated blood NH3 concentrations than are normal rats. These rats will develop behavioral encephalo- pathy at blood NH3 concentrations which would cause very little or no behavioral change in normal animals. Biochemical analyses of the brains of these rats have given insight into the pathogenesis of this encephalopathy. In this research, an impairment of the malate-aspartate shuttle was found, resulting in reduced cerebral ATP. Glutamate is used in the detoxification of NH3 and so is not available for its part in this cycle (Figure 3). Unlike previous results, brain a-ketoglutarate depletion did not occur when blood NH3 levels were elevated (Hindfelt et al., 1977). Ammonia Metabolism in Malnutrition In the process of digestion in a normal human gastrointestinal tract, approximately 4 grams of NH3 are produced during a 24-hour period. In addition to ingested nitrogenous nutrients, degradation of epithelial and bacterial debris contributes to the NH3 load. It has been estimated that 15 to 30% of urea produced in the liver is eventually secreted into the stomach or small intestine and hydrolyzed to NH . While the NH produced in the digestive tract is primarily 3 3 due to the action of bacterial ureases in the colon, some is produced 13y mucosal urease in the stomach and small intestine. Ammonia itself xnay be secreted into the stomach, the diffusion regulated by differences knetween blood and stomach NH3 concentrations. Once in the gastric lumen, the acidity of the contents ionizes the molecule and traps it + as 'the poorly diffusible NH4 ion. Ammonia may also be secreted by the jejtnnnn, thus reducing blood NH3 concentration. This is believed to be Ein active process in the jejunum, while NH3 absorption is passive. 24 manusnm oumunmmmmlmumamfi co oficofifim mo uommmm N2 \ 22:83.0 2L O x 28.592 ||v 28.33.. l \ \ 222352815 lllVoSESBe—ox O — a SaESEG \ .cfimun ecu :A .m enough 25:23.0 6:2 “-_._—37 SgooflaxoTl 2222 2222 TI 223235 D 3 £22 +a 1 minute. The resin was then allowed to settle and the supernatant aspirated. The wash with 3 m1 NH3-free H20 was repeated twice more or until the supernatant was free of foam to assure removal of plasma proteins. Final volume was then adjusted to 1.5 ml with NH3-free H20. At this time plasma NH3 was adsorbed to the resin. One milliliter of 0.1 N NaOH was then added to each tube and the tubes vortexed for 1 minute. After the resin was allowed to settle, 1 ml of the result- ing supernatant was removed to a NH -free 7 to 10 m1 glass test tube. 3 This supernatant contained NH3 which had been removed from the resin . . . + . by competitive adsorption of Na ions. To the tube containing the 1 ml of supernatant, 2 m1 of the phenol color reagent was added, followed by 2 m1 of alkaline hypochlorite. m . . . . . Tubes were then capped and mixed by inverSion. After miXing, tubes 1Vortex—Genie Mixer, American Hospital Supply Corporation, McGraw Park, IL. mParafilm "M" Laboratory Film, American Can Company, Greenwich, CT. 45 were placed in a 37 C water bath for 15 minutes. Tubes were then removed from the bath and mixed by inversion, and the solution was placed in 12 x 75 mm glass cuvettes. All tubes were then read after zeroing with an NH3—free H20 blank on a spectrophotometern set at 640 mu wave- length. Optical density (OD) of standards and samples was corrected (OD-C) by subtracting the OD of the NH3-free H20 blank which had been run through the resin procedure. Concentration of NH3 was calculated in HQ per 100 ml by the following formula: concentration standard x OD-C unknown OD-C standard concentration unknown = So that the benefit of all standards could be gained, the concentration of a given standard was divided by its OD-C. The result was called the F value of that standard. The PS of all standards were averaged and the resulting factor was multiplied by the OD-C of the unknown to obtain its concentration. Unless otherwise indicated, the above pro- cedure was used on all samples to be discussed. Recoveries and Serum vs Plasma All recoveries were run on plasma samples from the same human subject (FK). The procedure used was the same as for all other samples. Identical plasma samples were placed in 4 resin tubes, 0.5 ml each. To the first tube, nothing was added, while to the other 3 tubes, 0.5 ml of each of the previously prepared working standards was added. Ammonia concentrations were then determined as usual. The value obtained from the sample to which no standard was added was subtracted nColeman 44 Linear Absorbance Spectrophotometer, Perkin—Elmer Corporation, Oak Brook, IL. 46 from the NH3 values of the other 3 tubes. The result was the recovered concentration of NH3. When this concentration was divided by the known concentration of the standard added, a percent recovery was calculated. In order to determine the effect of clotting on blood NH3 values, 2 samples of blood were collected from each human subject at the same time. One sample was collected into a tube containing no anticoagu— lant, while the second tube contained disodium EDTA. Both tubes were placed in a 4 C refrigerator until the former sample had coagulated. Therefore, the same time elapsed between drawing of blood and addition to resin for both samples. After centrifugation, 0.5 ml of either serum or plasma was added to resin and NH3 determination was carried out on each as usual. Samples of plasma were saved from some days' experiments so that the effect of holding blood at -20 C could be seen. Tubes containing plasma were capped and placed in a -20 C freezer for 24 hours, 48 hours or 5 days. After the period of time had elapsed, the plasma samples were thawed and plasma NH was determined as previously described. 3 Vitamin E Deficiency in Rats Sixty-day-old Sprague—Dawley rats were divided into 4 groups. Group 1 was fed a control diet of corn and soybean meal with a trace mineral and vitamin mixture added. Group 2 was given the same diet with no vitamin E and .05 Ppm selenium in the vitamin-mineral mixture. Group 3 was given the vitamin E and selenium deficient diet with 0.15% silver acetate added to the drinking H 0. Group 4 was given a vitamin 2 E and selenium deficient diet, 0.15% silver acetate in the drinking H20 and 5% cod liver oil in the diet. After 5 months on these diets, 47 rats were anesthetized with ether and blood samples were drawn by cardiac puncture. The rats were then exsanguinated when the right ventricle was cut. Tissues were collected in 10% buffered formalin for histological examination. One rat from Group 3 and 2 rats from Group 4 were perfused with Karnovsky's fixative so that electron microscopic study of the liver could be done. This was accomplished by opening the chest of the anesthetized rat, infusing Karnovsky's fixative into the left ventricle, and cutting the right ventricle to allow escape of blood. The rats were thought to be fully perfused when limbs went into tonic extension. Tissue was then collected, diced into approximately 1 mm cubes, and placed in Karnovsky's fixative. Bovine Starvation Samples of blood for plasma NH3 determination were collected from 5 adult Holstein COWS which were used on a starvation experiment, These cows were then given no food for 6 days, while H O was provided 2 ad libitum. After the 6-day starvation, blood samples were again collected and plasma NH3 levels determined. The cows were then given either a 23% protein diet (hay, soybean meal, salt, dicalcium phosphate) or a 12% protein diet (hay, corn, salt, dicalcium phosphate). One cow given a low protein diet and 1 given a high protein diet had blood samples drawn 5 hours after resuming eating. Because the cows were to be subjected to surgery for liver biopsy, a week interval was allowed to pass with the cows being given either a high or a low protein diet before the next blood NH samples were collected. 3 Dietary Urea Blood NH3 levels were determined on 10 calves before they were started on dietary urea. The calves were begun on 0.2% urea at the 48 age of 2 to 4 months. At weekly intervals, the urea was increased to 0.4, 0.7 and 1.0% and held at 1.0%. Blood NH3 determinations were run at the same time weekly before the amount of dietary urea was raised. Blood urea nitrogen0 was also determined for the sixth through the ninth week of the experiment. On the thirteenth week of the experi- ment, blood samples for NH were collected before the morning feeding 3 and 1/2 hour after feeding. The calves' diet was prepared as follows: soy protein 1500 gm cerelose 5750 gm mineral mix 500 gm urea 20-100 gm CaCO3 100 gm vitamin mix 100 gm cellulose 150 gm lard 500 gm selenium mix 10 gm flora 5 gm terramycin 1 gm Clinical Cases Blood samples were collected from patients into sodium heparin tubes for determination of whole blood NH3 at the Veterinary Clinical Center (VCC) clinical pathology laboratory, Michigan State University. Remaining blood was spun down and plasma was used for NH3 determination using the cation exchange resin technique. The VCC laboratory used the technique of deproteinization with sodium tungstate and sulfuric acid as previously described (McCullough, 1967). Comparisons were made between results of the 2 techniques in both resting NH3 and NH3 tolerance tests. Histopathologic Technique Methods of fixation of tissue from rats has been described. Tissues for light microscopic examination were embedded in paraffin oUrea Nitrogen Colorimetric Determination Kit, Sigma Chemical Company, St. Louis, MO. 49 and stained with hemaxotylin and eosin (Luna, 1968). Tissue for electron microscopic study was washed in cacodylate buffer with 4.5% sucrose and then osmicated in osmium tetroxide. After dehydration, the tissue was embedded in Epon. Thick sections were stained with toluidine blue. Thin sections cut with a LKB ultramicrotome were stained with uranyl acetate and lead citrate and examined with a Phillips 300 electron microscope operating at 60 KV. Standard photo- graphic technique was used to process the negatives. RESULTS Blood Ammonia Determination Technique The cation exchange resin technique for blood NH3 determination was found to be rapid and convenient for use in both solitary clinical cases and experimental procedures in which several samples were to be run at once. Recoveries of NH3 standards from human plasma ranged from 83.1% to 96.0% on 3 separate sets of recoveries (Table 1). When plasma and serum samples were run on the same individual's blood, a highly significant difference (p<.0001) was found in NH3 results. The mean value of plasma NH3 was 15.24 ug/lOO m1, while the mean serum NH3 level was 56.94 ug/100 ml (Table 2). Holding plasma samples at -20 C did not have a significant effect on blood NH3 values after 24 hours, the mean only increasing from 19.49 ug/lOO ml to 19.72 ug/lOO ml. After 48 hours under the same conditions, however, a sig- nificant rise (p<.005) in plasma NH was seen, the mean rising from 3 18.50 ug/lOO ml to 31.92 ug/100 ml. Freezing the plasma samples for 5 days gave results not unlike those seen at 48 hours, the mean NH3 values rising from 16.18 ug/loo ml to 28.70 ug/lOO ml (Table 3). Vitamin E Deficient Rats Plasma NH3 did not rise significantly from controls in either rats given a vitamin E deficient diet or those given a vitamin E deficient diet with silver acetate in the drinking water. The means 50 51 Table 1. Recoveries of ammonia standards added to human plasma Recovered Recovered Sample OD OD—C F ug/lOO ml ug/lOO ml % First Recovery Blank .078 100 Ug/lOO m1 .139 .061 1639 200 Ug/lOO m1 .197 .119 1681 300 Ug/lOO m1 .253 .175 1714 FK .087 .009 av.l678 15.1 FK+100 Hg/100 m1 .137 .059 99.0 83.9 83.9 FK+200 Ug/loo m1 .193 .115 193.0 177.9 89.0 FK+3OO Ug/100 m1 .241 .163 273.5 258.4 86.1 Second Recovery Blank .078 100 Ug/lOO m1 .130 .052 1923 200 Ug/loo ml .187 .109 1835 300 Ug/lOO ml .242 .164 1829 FK .082 .004 av.l862 7.4 FK+100 Ug/100 ml .129 .051 95.0 87.6 87.6 FK+200 Hg/lOO m1 .182 .104 193.6 186.2 93.1 FK+300 Ug/lOO m1 .222 .144 268.1 260.7 86.9 Third Recovery Blank .079 100 Ug/lOO ml .130 .051 1961 200 Hg/lOO m1 .190 .111 1802 300 Ug/lOO m1 .248 .169 1775 FK .088 .009 av.1846 16.6 FK+100 ug/100 ml .133 .054 99.7 83.1 83.1 FK+200 ug/lOO ml .185 .106 195.7 179.1 89.6 FK+300 Ug/lOO m1 .244 .165 288.0 96.0 304.6 OD: optical density OD-C: concentration of standard F: OD-C of standard av: average of F values optical density corrected (OD of sample - OD of NH3-free H20 blank) 52 Table 2. Ammonia levels in simultaneously run human plasma and serum samples Plasma NH3 Serum NH3 Serum NH -Plasma NH Sample (pg/100 ml) (Hg/100 m1) (Hg/I00 ml) 3 PK 9.00 43.00 34.0 LS 8.00 54.10 46.1 MT 26.00 82.20 56.2 TM 17.60 44.90 27.3 JR 15.60 60.50 44.9 TE 15.24 56.94 sea 3.25 7.05 aStandard error X: mean 53 Table 3. Changes in ammonia levels in plasma samples held at -20 C Bovine Plasma NH3 Canine Plasma NH Human Plasma NH3 (pg/100 ml) (Hg/100 ml) (ug/IOO ml) Day 0 Day 1 Day 0 Day 2 Day 0 Day 5 29.20 33.80 16.90 24.80 13.30 20.30 15.60 24.40 13.20 26.60 11.60 27.00 27.30 9.40 18.80 39.00 8.30 27.00 21.40 15.00 22.60 40.80 31.50 40.50 9.80 24.40 18.80 40.80 19.50 11.30 20.70 19.50 9.80 20.70 21.40 24.40 11.70 11.30 29.20 22.50 32' 19.49 19.72 18.50 31.92 16.18 28.70 SE 2.40 2.46 1.32 3.83 5.21 4.24 a Standard error X: mean 54 of plasma NH in Groups 1, 2 and 3 were 54.27 ug/100 m1, 62.28 ug/100 ml, 3 and 65.67 ug/lOO ml, respectively. In Group 4, however, a highly significant (p<.005) rise in plasma NH was seen, the mean being 3 174.01 ug/lOO ml (Table 4). These rats had been given a vitamin E deficient diet, silver acetate in the drinking water, and 5% dietary cod liver oil. 0n light microscopic examination, livers of these rats had moderate diffuse hepatocyte vacuolization, which had not been apparent in livers of rats from the other groups (Figure 4). These vacuoles stained positively with oil red 0, indicating that they represent fatty change of hepatocytes. Electron microscopic examina- tion was made of the liver of 1 rat each from Groups 3 and 4. The most dramatic change was seen in the Group 4 rat, which had extensive proliferation of disorganized smooth endoplasmic reticulum (Figure 5). Bovine Starvation The changes in the cows' plasma NH levels after 6 days of star- 3 vation were so varied it was not believed that statistical significance could be drawn from the results (Table 5). 0f the S cows used in the experiment, 3 (7, 12, 17) had a reduction in plasma NH after 6 days 3 of starvation. The remaining 2 cows (9, 10) had elevated plasma NH3 as compared to pre—starvation levels. Of the 3 cows in the former group, 2 were started on low protein diets after starvation (7, 12). While plasma NH did rise above post-starvation levels in these two, 3 they did not reach the NH3 concentrations seen before starvation. Cow 17, however, was placed on a high protein diet and subsequently had a dramatic rise in plasma NH This plasma NH reached a level of 3° 3 approximately 3 times pre-starvation NH and almost 8 times post- 3 starvation NH3 concentrations. Two of the cows (9, 10) had an elevation 55 Table 4. Plasma ammonia levels in Sprague-Dawley rats with experi- mentally induced liver damage Group la Group 2b Group 3C Group 4d (mg/100 ml) (us/100 ml) (pg/100 ml) (ug/loo ml) 100.70 54.60 101.30 206.40 47.00 92.90 37.50 187.60 38.60 41.90 58.20 75.00 52.00 85.60 300.20 47.00 36.40 135.10 40.30 93.80 234.50 159.50 §’ 54.27 62.28 65.67 174.01 SEe 9.50 11.44 18.82 26.30 aGroup l = control bGroup 2 = vitamin E deficient CGroup 3 = vitamin E deficient with silver acetate dGroup 4 = vitamin E deficient with silver acetate and cod liver oil e Standard error X: mean 56 Figure 4. Light micrographs of liver of a rat given a vitamin E deficient diet, silver acetate, and cod liver oil. (a) Fine vacuolization of cytoplasm is present in periportal hepatocytes (arrows). X40. (Epon, toluidine blue) (b) Large fat vacuoles are present in some hepatocytes (arrows). X40. (Epon, toluidine blue) Figure 5. Electron micrographs of liver of a rat given a vitamin E deficient diet, silver acetate, and cod liver oil. (a) Note fat vacuoles in cytoplasm (arrow). X15,150. (b) Note the large amount of disorganized smooth endoplasmic reticulum (arrows). X27,450. 58 Table 5. Effect of 6 days of starvation on bovine plasma ammonia levels Cow Before Starvation After Starvation After Feeding Resumed No. (pg/100 ml) (HQ/100 m1)avb (ug/100 ml) 5 hr 1 wk 7 56.30 10.80 14.00 9 7.80 13.40 19.10 7.70 12 36.50 17.90 28.70 10 14.40 32.30 62.70 52.70 17 64.10 25.40 201.40 '§ 35.82 19.96 60.90 SEC 11.08 3.95 35.91 a . . . . After starvation, 7, 9 and 12 were given 12% prote1n diet b . . . . After starvation, 10 and 17 were given 23% prote1n diet c Standard error X: mean 59 of plasma NH after 6 days of starvation. The one fed a 12% pro— 3 tein diet (9) had a further rise in plasma NH 5 hours after beginning 3 the diet, while the NH dropped to approximately pre-starvation level 3 1 week later. The second cow (10) had an approximate doubling of the post-starvation NH 5 hours after beginning the 23% protein diet. 3 One week later, after continuing the high protein diet, the plasma NH3 was still high as compared with both pre- and post-starvation values. Dietary_prea When calveswemtebegun on 0.2% dietary urea, there was only a mild elevation in their plasma NH concentrations (Table 6), the 3 mean increasing from 28.79 ug/100 ml to 32.67 ug/lOO ml. Contrary to expectations, however, a highly significant drop (.0001