SEYMOUR SABESIN§, AND JOHN F. GRIFFITH*. Departments of *Pediatrics, §Medicine, I Pharmacology, and tNeurology and the tAnimal Resources Division, ...
Proc. Natl. Acad. Sci. USA Vol. 81, pp. 6169-6173, October 1984 Medical Sciences
Development of encephalopathic features similar to Reye syndrome in rabbits (animal model/encephalopathy/polyunsaturated fatty acids)
ELLEN S. KANG*, GERALD OLSONt, J. T. JABBOURf, SOLOMON S. SOLOMON§¶, MURRAY HEIMBERG§1', SEYMOUR SABESIN§, AND JOHN F. GRIFFITH* Departments of *Pediatrics, §Medicine, I Pharmacology, and tNeurology and the tAnimal Resources Division, University of Tennessee Center for the Health Sciences, University of Tennessee, Memphis, TN 38163; and the ¶Research and Medicine Services of the Veterans Administration Hospital, Memphis, TN 38104
Communicated by Ralph T. Holman, June 8, 1984
The progression of neurological abnormaliABSTRACT ties through four or five clinically distinguishable levels of deepening coma and the development of a fatty liver are the hallmarks of Reye syndrome. A number of animal models have been described that result in fatty liver formation with minimal, static, or catastrophic neurological changes. In this study, we attempted to produce neurological features in rabbits that reflected a rostral-caudal progression of abnormalities that could be categorized into clinically distinguishable levels reminiscent of Reye syndrome. This was accomplished by the intracisternal administration of 0.5-25 mg of 11,14-icosadienoic acid (20:2wo6) suspended in a mixture of rabbit serum and isotonic saline solution. A reproducible, dose-titratable spectrum of at least four levels of deepening coma could be produced at will. Increases in serum glutamate-oxaloacetate transaminase and creatine kinase and changes in serum glucose resulted 1-2 hr after the neurological abnormalities were evoked. Other unsaturated fatty acids produced similar responses. Those tested included 18:1w9, 18:2w6, 18:3a3, 20:3w6, 20:4o6, and 22:4w6 fatty acids. Saturated fatty acids, including 6:0, 8:0, 16:0, 18:0, and 20:0, failed to elicit these effects. The abnormalities were sustained for 30-120 min after a single dose. Full recovery was observed in some animals that had not reached the fourth level of our grading system for coma. Pretreatment of the rabbits with aspirin modulated the neurological abnormalities. Twenty micrograms of bee venom melittin, which activates endogenous phospholipase A2, administered intracisternally into rabbits also produced signs of level 3 (our grading system) coma for several hours. These findings suggest a possible role for polyunsaturated fatty acids in the development of Reye syndrome and offer a means of producing the neurological components of that syndrome in a laboratory animal.
Reye syndrome is a rare but clinically important childhood encephalopathy of unknown origin associated with fatty visceral changes and complex metabolic abnormalities (1, 2). The encephalopathy is a progressive one with four or five stages "each correlating with progressive rostral-caudal central nervous system involvement" as first recognized by Lovejoy et al. (3) and subsequently corroborated by others
(4-6).
Several animal models have been proposed, including the
use of a specific inhibitor of fatty acid oxidation (7), the infusion of those metabolites that are found in increased concen-
of liver dysfunction have been observed in each of these models. However, from the neurological point of view, only minimal, nonprogressive changes or, alternatively, collapse and death have been reported. In some instances, the infusion of glucose reversed or prevented these neurological abnormalities (7). Such features are in contrast to the progression of the encephalopathy of Reye syndrome through four (14) or five (3) distinguishable levels of deepening coma that cannot be reversed by the infusion of glucose alone. This study was, therefore, designed to produce neurologicalfeatures in rabbits which resemble those of Reye syndrome. Our approach was drawn from the report of Ogburn et al. (15), which indicated that patients with Reye syndrome had high levels of long-chain polyunsaturated fatty acids (PUFA) in their sera. These levels remained elevated in nonsurvivors despite exchange transfusion, whereas in survivors the PUFA levels could be reduced by blood exchange. We have found that direct injection of these compounds into the cisterna magna of rabbits produced an encephalopathy that could be graded through several distinct levels of neurological dysfunction that parallel the clinical disorder.
METHODS New Zealand White male rabbits (Oryctolagus cuniculus) weighing 2-4 kg were supplied by a local breeder. These animals were fed standard laboratory rabbit food in pellet form. In all, more than a hundred individual animals were used in this study. Serum measurements were obtained selectively from groups of comparable animals. Animals subjected to intracisternal punctures were anesthetized by induction with halothane/02 and maintained on a 3% mixture delivered through a nose cone. Skin over the occiput and foramen magnum was shaved. The animal was placed in a prone position on a table with the head flexed over the edge at right angles to the spinal column. A 25gauge 5/-in (16 mm) needle was used for fully mature rabbits. The needle was inserted through the foramen magnum, penetrating the dura mater into the cisterna magna. As soon as cerebrospinal fluid filled the needle, indicating successful placement into the cisterna magna, 0.1-0.2 ml of the suspension to be tested was injected using a tuberculin syringe atAbbreviations: SGOT, serum glutamate-oxaloacetate transaminase (L-aspartate:2-oxoglutarate aminotransferase, EC 2.6.1.1); SGPT, serum glutamate-pyruvate transaminase (L-alanine:2-oxoglutarate aminotransferase, EC 2.6.1.2); LDH, lactate dehydrogenase (L-lactate:NAD+ oxidoreductase, EC 1.1.1.27); CK, creatine kinase (ATP:creatine N-phosphotransferase, EC 2.7.3.2); PUFA, polyunsaturated fatty acids. The abbreviated w nomenclature for polyunsaturated fatty acids consists of the number of carbon atoms in the fatty acid, colon, the number of double bonds, w, and the number of carbon atoms beyond the last double bond including the terminal (w)
trations in the blood of affected patients (8, 9), and the inoculation of animals with viruses, either alone or in conjunction with special diets or prior exposure of the animals to chemicals (10-13). Fatty liver formation and biochemical evidence
The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. §1734 solely to indicate this fact.
carbon atom. 6169
6170
Medical Sciences: Kang et aLPProc. NatL Acad ScL USA 81 (1984)
tached to the needle. The needle was promptly removed. The animal was placed in the Trendelenberg (head-down) position for 1 min, then placed on a flat surface for observation. Animal behavior was observed, described, and graded by consensus of a veterinarian, a pediatric neurologist, and a pediatrician. Fatty acids used for intracisternal injection were purchased from Sigma or P-L Biochemicals as either the free acids or the sodium salts. All were claimed to be >99% pure by gas/liquid chromatography or thin-layer chromatography. Aliquots of free PUFA with 18 or more carbons were stored in sealed, N2-flushed vials until use. These fatty acids were not checked for possible oxidation. Fatty acids for injection were suspended in normal rabbit serum/isotonic saline (1:9), a solution estimated to contain 600-1200,g of protein in the volumes used. The principal fatty acid used was 11,14-icosadienoic acid (20:w6) because this compound is far less expensive than either 20:4w6 or 22:4w6. Other fatty acids tested included hexanoic acid (6:0), octanoic acid (8:0), palmitic acid (16:0), stearic acid (18:0), oleic acid (18:1w9), linoleic acid (18:2w6), linolenic acid (18:3w3), arachidic (icosanoic) acid (20:0), 8,11,14-icosatrienoic acid (20:3w6), arachidonic acid (20:4w6), and 7,10,13,16-docosatetraenoic acid (22:4w6). Melittin, the bee venom peptide known to activate membrane phospholipase A2, was purchased from Sigma. Mannitol was obtained from Lypho-Med as a sterile 25% solution (1.6 osM). Standard automated methods were used to measure serum glutamate-oxaloacetate transaminase (SGOT), glutamatepyruvate transaminase (SGPT), lactate dehydrogenase (LDH), creatine kinase (CK), bilirubin, triglyceride, urea nitrogen, creatinine, glucose, electrolytes, and ammonia. Only relevant changes in the serum constituents are presented. Aspirin and acetaminophen (University Hospital Pharmacy) were given by gavage as suspensions in water for 2.5 days before fatty acid injection. The doses, given in 3 aliquots per day, were 65 mg/kg and 30 mg/kg of body weight each day for aspirin and acetaminophen, respectively. Control animals were given an equal volume of water 3 times a day.
RESULTS Neurological Abnormalities Evoked by the Intracisternal Administration of the PUFA, 11,14-Icosadienoic Acid (20:2w6). Dramatic neurological abnormalities were reproducibly evoked in young and mature rabbits upon the introduction of 11,14-icosadienoic acid into the cisterna magna. The neurological abnormalities produced covered a wide spectrum of signs, which could be graded into four or five levels of severity (Fig. 1). Level 1: Within 2-5 min of withdrawal of anesthesia, hyperventilation develops. The rabbit is slower to recover from anesthesia than control animals. Spontaneous movements and response to tactile stimuli are diminished. Postural balance, neck support, righting reflexes, and muscle tone are normal. The animal is lethargic. Recovery occurs in 30-60 min. Level 2: The rabbit proceeds through level 1. By 3-5 min, respiration is markedly increased in rate and depth. Hyperactive panicky behavior, screeching, purposeless running, increased muscle tone, intermittent neck extension, dilated pupils, and occasional anisocoria are seen. The panicky behavior can lead to thrashing, jumping, or rolling over repeatedly. Tonic-clonic movements of the extremities are seen, as is lateral nystagmus. Righting reflexes are diminished and intermittent somnolence occurs. Level 3: Animals proceed through levels 1 and 2 within 10 min. Marked hyperventilation occurs, and more persistent opisthotonus and hind limb extension are seen than in level 2. Generalized convulsions occur with grimacing, clenching of the teeth, coarse tremors, jerking of the extremities, and bladder incontinence. Coma follows with cessation of vocalization. The rabbit may become immobile and unresponsive to tactile stimuli. Level 4: The rabbit proceeds through levels 1, 2, and 3 within minutes. All signs of level 3 are exaggerated with development of opisthotonus and convulsions that may lead to the next level, apnea. Level 5: As for level 4, but proceeding to apnea. Apnea developed while the heart beat was present in each fatal case. These levels of neurological abnormalities correlated strongly (r - 0.79) with the dose of the 20:2co6 fatty acid administered. Except for the rapidity of the progression of these signs, the clinical features are reminiscent of the pro-
FIG. 1. Some of the more characteristic postures assumed by animals in this study. C, control. The Roman numerals refer to the levels of encephalopathy described in the text.
Medical Sciences: Kang et aL gressive cerebral and brainstem dysfunction described by McNealy and Plum (16) and by Plum and Posner (17), which follows an orderly rostral-caudal loss of diencephalic, midbrain, pontine, and medullary function when the brainstem is involved, with supratentorial mass lesions (16, 17). The control injection of the serum/saline used to suspend the PUFA did not result in any detectable neurological abnormality. Recovery from halothane/02 was rapid, with the rabbit assuming a crouched position with normal tone and becoming fully alert and responsive to stimuli 2-3 min after removal of the anesthesia nose cone. Administration of the saturated fatty acid icosanoic acid (20:0) also failed to elicit abnormal responses. Thus, carbon chain length alone could not be important in evoking these abnormalities. A number of other unsaturated fatty acids were administered including 18:1w9, 18:2w6, 18:3c3, 20:3co6, 20:4w6, and 22:4cw6. All elicited the neurological responses. Arachidonic acid (20:4w6) and 7,10,13,16-docosatetraenoic acid (22:4w6) were more toxic; lower doses evoked more severe abnormalities. The saturated fatty acids 6:0, 8:0, 16:0, 18:0, and 20:0, as indicated above, did not precipitate these signs. (Higher doses were not tested.) The systemic administration of 4-8 times the lethal intracisternal dose of the 20:2w6, 20:4w6, and 22:4w6 fatty acids did not result in detectable abnormalities other than several minutes of mild hyperventilation. In preliminary experiments, pretreatment of the animals with aspirin for 96 hr had appeared to reduce the severity of the neurotoxic effects of icosadienoic acid (20:2CI6), whereas acetaminophen did not appear to alter the response to the PUFA. Because these studies had been conducted on a number of different days, an experiment was done in which antipyretic treatments and cisternal injections were completed simultaneously on all animals (Fig. 2). Again, the modulating effect of aspirin was observed at the various doses of PUFA
given.
Serum Changes. Elevations of SGOT and serum CK occurred in the PUFA-treated rabbits (Table 1). Serum glucose levels were not significantly different, although two of the nine treated animals had levels within the hypoglycemic range (44 and 54 mg/dl). The systemic administration of the PUFA through the marginal ear vein at doses 4 times the lethal intracisternal dose did not appreciably alter these values except for CK (SGOT, 40.4 units/ml; CK, 1370 units/ ml; urea, 17.0 mg/dl; glucose, 153 mg/dl). Pretreatment of animals with aspirin before administration of the PUFA significantly reduced the blood glucose level. A similar trend was observed in the acetaminophen-pretreated animals, but the difference was not statistically significant. Effects of Acute Hyperosmolar Iijection into the Internal Carotid Artery. To explore the possibility that the fatty acids could traverse the blood-brain barrier from the systemic circulation under conditions of hyperosmnolarity, which are known to occur in many Reye syndrome patients (18), the internal carotid artery was cannulated and infused with 1 g of mannitol followed by a bolus dose of 100 mg of the 20:2w6
6171
Proc. NatL Acad Sci USA 81 (1984)
Cc 0.
5
A
0
4
i
;
3 0
4U
* 21& 1
0
0
0
0
S 0
0
0
0 0
0
.
.
1
2
.
.
.
6 4 3 5 Intracisternal PUFA, mg
8
10 20 30
FIG. 2. Level of encephalopathy (our grading system) plotted as function of intracisternal dose of 20:2w6. Response of animals not pretreated with antipyretics (e). Response of animals pretreated with aspirin (o) or acetaminophen (A) given as suspensions in water by gavage for 2.5 days before study. Doses were 65 mg/kg over 24 hr for aspirin and 30 mg/kg over 24 hr for acetaminophen, given in three aliquots each day. a
acid suspended in serum/saline. Within minutes, the rabbits (two) exhibited all of the features seen in animals given 10 mg of icosadienoic acid via cisternal puncture. Mannitol alone (one rabbit), mannitol followed by the injection of serum/saline (one rabbit), and protein suspended in serum/saline (one rabbit) did not elicit these effects. Melittin Effects. Melittin, a specific bee venom factor that activates phospholipase A2, was infused systemically or injected into the cisterna magna of 10 rabbits to stimulate hydrolysis of glycerol esters of polyunsaturated long-chain fatty acids. Systemic administration of 10 mg of melittin was lethal within minutes, whereas 1 mg resulted in respiratory difficulties (dyspnea was the most striking feature) and prominence of the eyes, all of which resolved spontaneously within 20 min. The injection of 0.1-10 mg of melittin into the cisterna magna resulted in the rapid onset of the same symptoms elicited by unsaturated fatty acids and led to death in every case. The number of rabbits given each intracisternal dose was as follows: 10 mg, 1; 2.5 mg, 1; 1.5 mg, 2; 1.0 mg, 1; 0.3 mg, 1; 0.25 mg, 1. All of these doses ended in rapid fatality, whereas 2 rabbits given 20 ,ug and 1 rabbit given 10 A.g showed a slower progression of symptoms. Neurological abnormalities similar to those described above were entered more slowly and were sustained for several hours. Thus, the symptoms evoked by the PUFA were reproduced by the introduction of 20 pg of melittin into the central nervous system compartment via the cisterna magna.
DISCUSSION The etiology of Reye syndrome is unknown. The syndrome is characterized by the acute onset of a rapidly progressive
Table 1. Levels of serum components Glucose, mg/dl Urea, mg/dl CPK, units/ml SGOT, units/ml Treatment ± 9.2 ± 159.8 23.5 1.4 77t 414 = 27.3 7.5* Untreated animals (n 5) 219 ± 41.5* (44-440) 2.5 22.3 1266 ± 128 58.1 222.1 20:2w6 (n = 9) 111.8 ± 21.3* 19.8 + 2.1 1032 ± 211 264.6 79.0 Aspirin, 20:2w6 (n = 6) 237.5 (144-331) 20.5 (17-24) 972 (445-1500) 39.7 (35 44.4) Aspirin (n = 2) 133.7 ± 12.9 17.8 ± 4.3 1160 ± 209 212.4 ± 67.3 Acetaminophen, 20:2w6 (n = 6) 166.7 ± 17.7 20.3 ± 5.9 1129 ± 318 48.4 ± 16.4 Acetaminophen (n = 3) The 20:2w6 fatty acid was injected intracisternally. Aspirin and acetaminophen were given by gavage over a 2.5-day period before fatty acid injection. Results are shown as mean ± SEM. Ranges are given in parentheses where appropriate. *P < 0.05. tp < 0.0005.
6172
Medical Sciences: Kang et aL
encephalopathy and hepatic steatosis after a prodromal ill-
ness with one of several different viruses (influenza B and varicella are most common; refs. 19 and 20). This syndrome is biphasic, with the onset of symptoms appearing several days after the peak of the prodromal illness (20, 21). There is no evidence of a tissue inflammatory response, viruses are rarely recovered from tissues, and cellular and humoral responses to antigens appear to be appropriate (22). Thus, an alteration in the metabolic response of the host to the viral insult or the postinfectious state must be the basis of Reye
syndrome. Many metabolic abnormalities have been reported in this disorder. Derangements in carbohydrate, lipid, and amino acid metabolism reflect an overall catabolic state as well as highly specific metabolic disturbances in the liver and adipose tissue (2, 23, 24, **). Recently, Ogburn et al. (15) reported elevated levels of free long-chain PUFA in the sera of 10 Reye syndrome patients and decreases in PUFA in the serum phospholipids. Increases as great as 10- to 25-fold were found for some of these fatty acids, especially the cw6 fatty acids. Eight of 11 patients were exchange transfused. In survivors, the concentration of these fatty acids dropped after exchange, whereas the levels rose in nonsurvivors. Therefore, we injected a variety of PUFA into the cisterna magna of rabbits. A spectrum of highly reproducible, dose-related neurological signs resulted. The array of neurological abnormalities was reminiscent of the rostral-caudal progression of cerebral and brainstem dysfunction described by Plum and his co-workers (16, 17). These signs were sustained for 30 min to hours, depending on the dose. Admittedly, the neurological abnormalities observed in man may not be exactly replicated in the rabbit, but striking parallels are obvious. The exact anatomical location of lesions that produce decerebrate and decorticate posturing have already been reported. In higher animals such as the cat, dog, and monkey, intercollicular section of the brainstem produces decerebrate posturing, and extirpation of the cerebral cortex and its underlying white matter to the level of the basal ganglia produces decorticate posturing (25). Temporary and reversible decerebrate posturing can also be produced in animals by cryogenic stereotactic lesions placed in the brainstem reticular formation (26). The progression of signs in our model (from lethargy to hyperactivity and agitation through abnormalities of muscle tone with decerebration, decortication, and finally, flaccidity and apnea) can be controlled through the dose of the PUFA administered and may be useful in studies of metabolic coma as well as of the consequences of expanding lesions of the central nervous system. Not all patients with Reye syndrome progress through all four or five stages of the disease. More patients are being recognized who enter only stage 1 of Reye syndrome (27) without further advancement of the disease. However, the risk of further progression in stage 1 patients has been recognized and estimated to be from 14% (28) to 34% (27). Ogburn et al. (15) postulated that the origin of the PUFA in the sera of Reye syndrome patients could be due to activation of phospholipase A2 in the liver or brain with cleavage of the PUFA from position 2 of the phospholipids. The effects of melittin that we observed support this hypothesis. The brain has a unique response to a variety of insults. Brain swelling or edema occurs after head trauma, ischemia, infection, toxic injury, etc. This response may depend on the local release of covalently bound PUFA from neuronal membranes as free PUFA (29). A lipid-soluble component of granulocyte membranes has been shown to induce cortical swelling in vitro and has been proposed as the agent respon-
**Arcinue, E. L., Reye's III Symposium, Nov. 6-7, 1980. Detroit, MI (abstr.).
Proc. NatL Acad Sci. USA 81
(1984)
sible for the massive edema associated with fatal cases of bacterial meningitis and brain abcess (30). This component has been specifically identified as the PUFA (31). Recently, Chan et al. (32) have demonstrated that, in rats, the intracerebral injection of PUFA (18:3 and 20:4), in contrast to saturated (9:0 and 16:0) and monounsaturated (18:1) fatty acids, caused significant increases in cerebral water and sodium content but decreases in potassium content and in the activity of the Na+/K+-dependent ATPase. However, no clinical abnormalities were discerned. On the basis of the 125I-labeled bovine serum albumin-space studies and the staining of the brain with Evans blue dye, Chan et al. concluded that vasogenic and cellular brain edema were induced by the intracerebral injection of PUFA. The absence of clinical abnormalities is of interest and could be due either to species differences or to sequestration of PUFA within the substance of the brain, whereas our mode of delivery of the PUFA allows the entire surface of the brain to be in contact with the PUFA. It is possible that the neurological features of Reye syndrome result from the release of PUFA within the central nervous system itself. There is ample evidence that these PUFA are present in normal mammalian brain tissue (3335). Chaves-Carballo and Ellefson (36) have, however, reported no differences in the brain cholesterol, triglyceride, or phospholipid content in Reye syndrome. It is unknown whether the concentrations of free fatty acids are elevated in the central nervous system of Reye syndrome patients. On the other hand, PUFA levels are known to be elevated in the blood of affected patients (15). If the blood-brain barrier were compromised, these fatty acids could gain access to the central nervous system compartment. The infusion of 1 g of mannitol into the internal carotid artery followed by the bolus injection of 100 mg of 11,14-icosadienoic acid resulted in the neurological symptoms characteristic of 10 mg of the same fatty acid injected directly into the cisterna. The infusion of mannitol followed by an infusion of protein solution was without effect. It is likely that the mannitol caused the tight junctions to be temporarily compromised (37-39) and allowed the fatty acid to traverse the blood-brain barrier. The dose of mannitol administered contributed 6.4 mosmol to the plasma compartment. In a 2- to 3-kg rabbit, this would result in an -60% increase in the serum osmolality. (The estimated plasma volume of the rabbit is 40 ml. Normal human plasma is 270 mosM. Forty milliliters of rabbit plasma contains 10.8 mosmol. The addition of 6.4 mosmol of mannitol would increase the total mosmol in 40 ml of plasma to 18.2 mosmol, a 64% increase.) Mannitol is frequently used in the clinical management of patients with Reye syndrome at doses of 1-2 g/kg of body weight as a bolus dose, followed by smaller doses as required for the osmotic reduction of cerebral edema (5). Since Ogburn et al. (15) found high levels of PUFA in the sera of Reye patients, and we observed that serum PUFA can apparently translocate across the bloodbrain barrier after a bolus dose of mannitol, mannitol, as well as other hypertonic agents, should be used judiciously. Since translocation appears to be temporary, the intermittent use of hypertonic agents only as indicated would be preferable to continuous use over a period of time. It is of interest that the rat brain has been shown to take up labeled fatty acids injected into the carotid artery and that the fatty acids are taken up as the free fatty acids (40). The mechanism whereby the unsaturated fatty acids produce the neurological response is unknown. The moderating effects of aspirin pretreatment suggest a role for the prostaglandins and their metabolic products, which could be modified by the PUFA introduced by us. Our model system does not parallel Reye syndrome in that the viral prodrome does not occur and that the liver is not grossly fatty. The latter could well be due to the brevity of
Medical Sciences: Kang et aL our observations. Alternatively, the neurological features produced by the PUFA may be unrelated to development of fatty liver. We thank Ms. Mary Capaci and Ms. BaIjeet Sawhney for technical assistance, Dr. Carolyn Blackwell for the laboratory analyses during our pilot effort, and Mrs. Myrna Osteen for manuscript preparation. This study was supported in part by Grants NIH HD 11657 and NS 16383 (E.S.K.) and short-term Research Training Grant AM 07405 from the United States Public Health Service, by Veterans Administration Grant MRS 8036 (S.S.S.), and by grants-in-aid from Connaught Laboratories and Plough, Inc. 1. Reye, R. D., Morgan, G. & Baral, J. (1963) Lancet li, 749-752. 2. Haymond, M. W., Karl, I. D., Keating, J. P. & De Vivo, D. C. (1978) Ann. Neurol. 3, 207-215. 3. Lovejoy, F. H., Smith, A. L., Bresnan, M. J., Wood, J. N., Victor, D. I. & Adams, P. C. (1974) Am. J. Dis. Child. 128, 36-41. 4. Bobo, R. C., Schubert, W. K., Partin, J. C. & Partin, J. S. (1975) J. Pediatr. 87, 881-886. 5. Dobrin, R. S. (1980) Res. Staff Physician 26, 79-93. 6. Sherard, E. S. & Cooper, R. F. (1974) in Reye's Syndrome, ed. Pollack, J. D. (Grune & Stratton, New York), pp. 27-37. 7. Glasgow, A. M. & Chase, H. P. (1975) Pediatr. Res. 9, 133138. 8. Trauner, D. A. & Adams, H. (1981) Pediatr. Res. 15, 10971099. 9. Colon, A. R., Ledesma, F., Pardo, V. S. & Sandberg, D. H. (1974) Am. J. Dig. Dis. 19, 1091-1101. 10. Henle, G. & Henle, W. (1944) Science 100, 410-411. 11. Crocker, J. F. S., Ozere, R. L., Rozzo, K. R., Digout, S. C. & Hutzinger, 0. (1974) Lancet fi, 22-24. 12. Hug, G., Bosken, J., Bove, K., Linneman, C. C., Jr., & McAdams, L. (1981) Lab. Invest. 45, 89-109. 13. Deshmukh, D. R., Maassab, H. F. & Mason, M. (1982) Proc. Natl. Acad. Sci. USA 79, 7557-7560. 14. Huttenlocher, P. R. (1972) J. Pediatr. 80, 845-850. 15. Ogburn, P. L., Jr., Sharp, H., Lloyd-Still, J. D., Johnson, S. B. & Holman, R. T. (1982) Proc. Natl. Acad. Sci. USA 79, 908-911. 16. McNealy, D. E. & Plum, F. (1962) Arch. Neurol. 7, 10-32. 17. Plum, F. & Posner, J. B. (1980) in The Diagnosis of Stupor and Coma, Contemporary Neurology Series (Davis, Philadelphia), 3rd Ed.
Proc. Natl. Acadl Sci. USA 81 (1984)
6173
18. Alvira, M. M. & Forman, D. T. (1974) Ann. Clin. Lab. Sci. 4, 477-483. 19. Corey, L., Rubin, R. J., Bregman, D. & Gregg, M. B. (1977) Pediatrics 68, 702-708. 20. Linneman, C. C., Jr., Shea, L., Parin, J. C., Schubert, W. K. & Schiff, G. M. (1975) Am. J. Epidemiol. 101, 517-526. 21. Hochberg, F. H., Nelson, K. & Janzen, W. (1975) J. Am. Med. Assoc. 231, 817-821. 22. Linneman, C. C., Jr., Shea, L., Kauffman, C. H., Schiff, G. M., Partin, J. C. & Schubert, W. K. (1974) Lancet ii, 179182. 23. Kang, E. S., Gates, R. E. & Wrenn, E. L., Jr. (1982) Biochem. Med. 27, 180-194. 24. Kang, E. S., Schwenzer, K., Wall, H. P., Jabbour, J. T., Shade, R., Crofton, J. T. & Share, L. (1982) Biochem. Med. 27, 121-134. 25. Denny-Brown, D. (1962) The Basal Ganglia and Their Relation to Disorders of Movement (Oxford Univ. Press, London). 26. Gilman, S. & Van Der Meulen, J. T. (1965) Arch. Neurol. 13, 297-306. 27. Lichtenstein, P. K., Heubi, J. T., Daughtery, C. C., Farrell, M. K., Sokol, R. J., Rothbaum, R. J., Suchy, F. J. & Balistreri, W. F. (1983) N. Engl. J. Med. 309, 133-139. 28. Partin, J. C., Partin, J. S. & Schubert, W. K. (1978) Pediatr. Res. 12, 555A (abstr.). 29. Fishman, R. A., Sligar, K. & Hake, R. B. (1977) Ann. Neurol. 2, 89-94. 30. Fishman, R. A. (1975) N. Engl. J. Med. 293, 706-711. 31. Chan, P. H. & Fishman, R. A. (1978) Science 201, 358-360. 32. Chan, P. H., Fishman, R. A., Caronna, J., Schmidley, J. W., Prioleau, G. & Lee, J. (1983) Ann. Neurol. 13, 625-632. 33. Koeppen, A. H., Barron, K. V. & Mitzen, E. J. (1971) Brain Res. 35, 199-214. 34. Mohrhauer, H. & Holman, R. T. (1963) J. Neurochem. 10, 523-530. 35. Sinclair, A. J. (1975) Proc. Nutr. Soc. 34, 287-291. 36. Chaves-Carballo, E. & Ellefson, R. D. (1980) Arch. Neurol. 37, 165-167. 37. Rapoport, S. I. (1976) Blood-Brain Barrier in Physiology and Medicine (Raven, New York), pp. 100-113. 38. Brightman, M. W., Hori, M., Rapoport, S. I., Reese, T. S. & Westergaard, E. (1973) J. Comp. Neurol. 152, 317-326. 39. Rapoport, S. I. (1979) Ciba Found. Symp. 56, 237-251. 40. Dhopeshwarkar, G. A., Subramanian, C., McConnell, D. H. & Mead, J. F. (1972) Biochim. Biophys. Acta 255, 572-579.
