Favism

SECOND EDITION TOXIC CONSTITUENTS OF PLANT FOODSTUFES | EDITED BY IRVIN E. LIENER Department of Biochemistry College of Biological Sciences University of Minnesota St. Paul, Minnesota A Series of Monographs A complete list of titles in this series appears at the end of this volume. A Subsidiary of Harcourt Brace Jovanovich, Publishers New York London Toronto Sydney San Francisco

CHAPTER 9 Favism | J. MAGER, M. CHEVION, AND G. GLASER I. Favism as an Inborn Error of Metabolism: Epidemiological, Genetic, and | | | | Enzymological Aspects of the Disease ....................,,................ A, Brief Description and Epidemiology of the Disease ........................ B. Geographic Distribution of Favism .............................,....... C. Role of Glucose-6-Phosphate Dehydrogenase Deficiency in the Etiology of Favism D. Ethnic Distribution of G6PD Deficiency and Favism ....................... E. Mode of Inheritance of G6PD Deficiency ..............,...........,..... F. Molecular Characteristics of Normal and Mutant G6PD .…… G. Correlation between the Degree of G6PD Deficiency and the Severity of Clinical Symptomatology …… … … neen H. Differences between Caucasian and Negro Types of G6PD Deficiency ........ I. Possible Role of Genetic Determinants Other Than G6PD Deficiency in the Pathogenesis of Favism......................................,....,... J. Methods of Detection of G6PD Deficiency ............................... II. The Selective Toxicity of Fava Beans: Search for the Causative Agent of Favism... A. Effect of Ingestion of Fava Beans on the Life Span of G6PD-Deficient Erythrocytes eeen B. Studies with Crude Fava Bean Extracts ...........,.........,............. 266 266 267 268 269 271 272 277 277 | i | | 273 274 | 274 275 277 C. Fractionation of Fava Bean Extracts ...,...............,.,.... DC. 279 I D. Structure and Properties of Vicine, Convicine, and Their Aglycones .......... E. Effects of Divicine and Isouramil on Red Cell Metabolism in Vitro: Synergistic 279 | i Interaction of Isouramil and Ascorbic Acid....................,....,...... F. Possible Etiological Role of Divicine and Isouramil in Favism............... 281 283 Ì | G. Effect of 3,4-Dihydroxyphenylalanine on Erythrocyte GSH and Critical Evaluation of Its Postulated Role in the Etiology of Favism ...........,,............. | 284 IM. The Mechanism of the Biochemical Lesion Underlying Red Cell Destruction in Drug-Induced Hemolysis and Favism ..................................,.... IV. Concluding Remarks ..................,,.........,.,.................... References... „nnn ANA | Î 285 288 | | 288 i | | | | 265 TOXIC CONSTITUENTS OF PLANT FOODSTUFFS, SECOND EDITION Copyright © 1980 by Academic Press, Inc. All rights of reproduction in any form reserved. ISBN 0-12-449960-0

1. MAGER, M. CHEVION, AND G. GLAS CAL, OF METABOLISM: EPIDEMIOLOGI |. FAVISM AS AN INBORN ERROR S ECT ASP L GICA MOLO GENETIC, AND ENZY OF THE DISEASE gy of the Disease A. Brief Description and Epidemiolo by the Italian physician L. Montano to The term ‘‘favism’’ was coined in 1894 (fava following ingestion of broad beans designate an acute hemolytic anemia Itis . 1958) al., ef faba plant (Sansone beans) or inhalation of pollen of the Vicia toxic of ence pres already recognized the possible that the ancient Greeks had ion ematician Pythagoras founded a relig math The constituents in broad beans. the g eatin of lness of souls and the sinfu based on the tenets of the transmigration descriptions, ron, 1973). The first authentic broad bean (Russell, 1965; Wald 1850s (see literature date back to the midhowever, of this disease in the medical and ly growing number of case reports Aurichio, 1935). Since then, a rapid ate separ a as s statu ibuted to establish its clinical studies of favism have contr . beans fava of the etiological role nosological* entity and to corroborate Italy reviewed by Fermi and Martinetti in sm favi of In a series of 1211 cases (1905), 725 were due to ingestion tion of pollen, whereas of broad beans, 459 were attributed to inhalained origin. In a the remaining 27 cases were of undeterm 4 (0.7%) were ey of 579 cases of favism in Iran, only recent epidemiological surv fava plant pollen, whereas consumption thought to be attributable to exposure to e for the remainder of the cases (Hedayat of broad beans was held to be responsibl confirm the occurrence of pollen-induced et al., 1971). Other reports failed to . Thus, Kattamis et al., 1969; Belsey, 1973) outbreaks of favism (Chung, 1965; lackstill is n polle of the causative role definitive and unequivocal evidence for be to d foun were s bean of dry or cooked ing. Hemolytic crises due to ingestion were ks attac e sever whereas the most usually of a rather moderate intensity, observed after eating of fresh raw fava seeds (Luisada, 1941). In rare instances, of Pisum sativum (garden peas) and plants other than Vicia faba, e.g., seeds as a ena hybrida, have been incriminated pollen of Anagyris foetida and Verb al., et za Lariz (see m ly similar to favis cause of a hemolytic syndrome close umed cons ously previ had iduals who 1960). Favism has been observed in indiv expofirst ts; others developed the disease on fava beans with no untoward effec . Recur1973) y, Belse Kattamis et al., 1969; sure (Angelova and Andrev, 1959; (Belsey, 1973). rent attacks of favism are not uncommon a seasonal incidence of the disease, Epidemiological studies in Italy revealed in l-May, when the plant blossoms, and characterized by two peaks: in Apri In . 1941) ada, (Luis et mark s appear on the July-August, when the fresh ripe bean al. (1969) found a pronounced seasonal ef so Dono ion, tigat a more recent inves 267 peak in the Caspian littoral sea region between the middle of May and the middl of June, whereas in Abadan the highest incidence of favism occurred in A 7 in both instances, the seasonal peaks coincided with the harvesting of fava = the respective regions. na Data accumulated over the past 10 years reveal that favism is much m common in children under the age of 10 than in adults, with the highest fre. quency observed in the age group of 2-4 years, probably resulting from the fi ii exposure of these infants to the beans. There are quite a few reports, howev È favism occurring during the first year of life, and even in breast-fed i A = apparently because of transmission of the noxious agent through the Sher milk (Angelov and Andrev, 1959; Chung, 1965; Kattamis 1969 197 D in ane. cease shows a marked predilection for the male sex the fle ‘female OT Fe EE: series studied between 21:1 and 2.7:1 (Hedayat et al, The clinical picture of favism is governed by the symptomatology inherent i the hemolytic event, the major manifestations being pallor, fatigue, dys n nausea, abdominal or back pain, fever, and chills. The hemalysiö is of uBEL intensity and gravity, the more severe cases being attended by hemo Jobin. a and jaundice (see Beutler, 1972) and occasionally by acute renal failure (Sch no and Fritz, 1968; Symvoulidis er al., 1972). The onset of hemolysis rita extremely rapid and abrupt, especially in cases attributed to pollen inhalation, with symptoms starting within a few minutes following the noxious exposure. I | the vast majority of cases, an interval of about 5-24 hr intervenes betwe the ingestion of the broad beans and the first manifestations of the hemolytic m ke drin in rare instances the onset of the symptoms may be delayed until the a ere om the consumption of broad beans (Luisada, 1941; . a of the disease is usually self-limited, the acute stage lasting 24-48 r, and is then followed by prompt, spontaneous recovery. In contrast to th usually benign outcome of the disease in adults, a 6-8% case fatality Eee a in children under 6 years of age (Fermi and Martinetti, 1905). More recentl owever, the mortality figures have been greatly reduced on account of the ad ni of blood transfusion therapy (Crosby, 1956; Hedayat et al., 1971) “en B. Geographic Distribution of Favism Favism exhibits a striking prevalence in the insular and littoral regi on i I arca and in the Middle East (Sardinia, Sicily, im an rn " an anne Greece, Rhodes, Cyprus, Turkey, Lebanon, Israel, Iraq pra, A u nn Islands, Algeria, Egypt, and Sudan) (Luisada, 1941; Hedayat n 7 ; Belsey, 1973; Amin-Zaki et al., 1972; Hassan, 1971). The disease also frequently encountered in China (Chung, 1965) and Bulgaria (Angelov of the classification *Nosology is defined as the science of diseases.

and Andrev, 1959). Sporadic cases have been reported from Germany (Gehrmann ef al., 1963, Johannsen ef al., 1968), France (Auquier et al., 1968), Poland (Rockicka-Milewska et al., 1968; RoZynkowa et al., 1970, 1971), Rumania (Schneer, 1968), Yugoslavia (Vince-Ribarié, 1962), and Singapore (Wong, 1972) (see also Table I). The highest incidence was observed on the island of Rhodes (Kattamis et al, 1969). The disease was found to be unevenly distributed on the island, with regions in which up to 40 cases per 1100 male inhabitants were reported. The overall prevalence of favism in the total population of the island during the years 1952-1965 was found to be 1.7-5.4 per thousand. A similarly high prevalence was recorded in Sardinia, with about five cases of favism per 1000 inhabitants (Crosby, 1956). This highly selective geographic distribution of favism is particularly puzzling since fava beans are grown and consumed almost all over the world as a cheap and popular staple food, distinguished by its relatively high content of carbohydrates and proteins (58 and 25%, respectively) per edible portion of mature dry seeds (see U.S. Department of Agriculture, 1963). C. Role of Glucose-6-Phosphate Dehydrogenase Deficiency in the Etiology of Favism The various concepts advanced in the earlier literature on the infectious (see Luisada, 1941), toxic (Gasbarrini, 1915), or immunological (Manai, 1929; Dacie, 1954) etiology of favism appear today to be of historic interest only. These views, apart from being devoid of solid experimental ground, failed to account for some of the most salient features of the disease, namely, its restricted geographic and ethnic distribution and its pronounced familial tendency (Luisada, 1941). The door to an understanding of the true pathogenetic nature of favism was opened by the elucidation of the nature of the inborn error of metabolism underlying the so-called drug sensitivity, i.e., an abnormal propensity of certain individuals to develop acute hemolysis in response to treatment with primaquine and a variety of other drugs (see Beutler, 1972). Shortly after the discovery by Dern et al. (1954) that ‘‘primaquine sensitivity’’ is determined by an intrinsic abnormality of the erythrocytes, it was revealed by Beutler ef al. (1955, 1957) that the susceptible red blood cells exhibit a relatively low content of reduced glutathione (GSH) and an enhanced rate of GSH destruction on incubation with 1-acetyl-2-phenylhydrazine in the presence of glucose. This so-called glutathione instability of drug-sensitive red blood cells was then shown by Carson et al. (1956) to be due to a deficiency of the nicotinamide adenine dinucleotide phosphate (NADP)-linked glucose-6-phosphate dehydrogenase (G6PD) and the resultant incapacity of these cells to maintain an adequate supply of NADPH to cope with the increased demand for GSSG (oxidized glutathione) reduction im- 9. FAVISM 269 posed by the challenging drugs (see Section II). In normal erythrocytes exposed to an oxidant stress, the enhanced rate of GSSG formation is compensated by a parallel increase in the rate of its reduction to GSH as a result of the concomitant stimulation of the G6PD activity, governed by an intrinsic regulatory mechanism (see Yoshida, 1973). The selective vulnerability of the red blood cells to this enzyme deficiency is accounted for by their critical dependence on the pentose phosphate shunt as the sole mechanism for NADPH generation, due to the lack of De mate pathways for NADPH supply that are present in other cells (Beutler, Crosby (1956), in a brief and brillant report concerned with the clinical and epidemiological aspects of favism in Sardinia, was the first to point out the resemblance of this disease to the primaquine-induced hemolytic anemia insofar as in both instances the red blood cells are capable of normal survival unless they are challenged by the noxious agent. He suggested, by analogy to primaquine sensitivity, that a hereditary enzymatic deficiency may be the underlying cause of the susceptibility to favism. The essential correctness of Crosby’s idea was soon substantiated by direct experimental evidence obtained independently by Sansone and Segni (1956 1957a,b, 1958) in Italy and by Szeinberg et al. (1957, 1958a,b) in Israel. These workers found that the GSH content of erythrocytes from persons known to have been affected by favism tends to be significantly lower (mean values below 50 mg %) than in normal individuals (mean range, 60-88 mg %) (Sansone and Segni, 1956; Szeinberg et al., 1957). More significantly, the red cell GSH level was found to decline sharply during the acute phase of favism. (Szeinberg and Chari-Bitron, 1957; Larizza et al., 1958), concomitant with the frequent appearance of methemoglobin and intracellular inclusions called Heinz bodies, similar to those observed in drug-induced hemolysis (Panizon and Pujatti, 1957; Larizza et al., 1960). Furthermore, in all persons with a past history of favism the erythrocyte GSH proved to be unstable in Beutler’s acetylphenylhydrazine test in vitro (Sansone and Segni, 1957a; Szeinberg et al., 1958a). Finally, and most important, the GSH instability was invariably associated with a pronounced G6PD deficiency of the red blood cells (Sansone and Segni, 1958; Szeinberg et al., 1958b; Larizza et al., 1958; Zinkham et al., 1958). | D. Ethnic Distribution of G6PD Deficiency and Favism G6PD deficiency is probably the most common genetically determined enzymatic defect in human beings, affecting, according to a rough estimate by Carson (1960), about 100 million people of all races throughout the world. Its geographic distribution closely parallels that of malaria, presumably because of the selective advantage offered by the enzymatic deficiency in increasing the resistance of the red blood cells to infestation by Plasmodium Jalciparum (Allison and

7 EAUISM TABLE I Clyde, 1961; Luzatto et al., 1969). The frequency of occurrence of this inborn error of metabolism is widely dissimilar in the different ethnic groups. As shown in Table I, the highest figures of incidence of G6PD deficiency have been INCIDENCE OF G6PD DEFICIENCY AND OCCURRENCE OF FAVISM IN DIFFERENT ETHNIC GROUPS Reference“ (%) = | Incidence Ethnic groups recorded, in decreasing order, in some of the oriental Jewish communities of Israel, Sardinians, Cypriot Greeks, American Negroes, and certain African popu- Favism G6PD Deficiency Occurrence? _ Reference lations. On the other hand, the abnormal trait is extremely rare or virtually absent in northern European nations and among Ashkenazic Jews (of European descent), North American Indians, and Eskimos (see Motulsky, 1960). It should be stated, however, that the available statistical data are on the whole rather incom- Soa po 13 1 plete and practically nonexistent in many underdeveloped areas lacking the Canada (Nova Scotia) 3-27 2 elementary facilities for the detection of G6PD deficiency. Congo (Kinshasa) 18-23 3 Pygmies 4 A No detailed statistics are available concerning the ethnic distribution of favism. As may be seen, however, from Table I, the occurrence of favism does Bantu Nigena ni 24 5 6 not parallel the frequency of the G6PD-deficient trait in the different populations. Most conspicuous in this respect is the complete absence of favism in North Gambia 15 Sudan 71-8. Ghana eng Asians ardinia sa age American Negroes (see Beutler, 1971). + 9 E. Mode of Inheritance of G6PD Deficiency + 10 x a (males . È : Sicily, ae soi Crete) 0.7-3 11 i Il G6PD deficiency is transmitted by a gene located in the X chromosome. This en Gal 7 Cyprus (Greeks) Ashkenazic Jews Oriental Jewish communities (males) Kurdish ran 22 7-11 0.2 12 13 + u rn | | | | Yemenite 5 mode of inheritance was deduced from extensive family studies (Childs et al., 1958; Szeinberg et al., 1958c; Larizza et al., 1960) and from the parallel segregation pattern observed when the G6PD-deficient trait and some other sexlinked anomaly, such as color blindness (Porter et al., 1962) or hemophilia A (Boyer and Graham, 1965), happened to coexist in the same individual (see also Aebi, 1967). Accordingly, the enzyme deficiency is fully expressed in the hemizygous (XY) male, because the mutant gene (X) is not counteracted by the North African 2 normal allele (X). On the other hand, full expression is rather uncommon in | females, since it depends on the statistically rare occurrence of a homogyzous | 14 16 A 15 15 17 | Israeli Arabs : Lebanon 26 19 + 19,26 mutant genotype (XX). In the majority of affected females the G6PD deficiency | rk 11 20 + E is found to be of a partial or intermediate nature, in accordance with the expected Iraq (males) | he D ca preponderance of the heterozygous (XX) constellation (Larizza ef al., 1960). | (adults “i io en 15-24 23 Iran 10 24 India 6 3 Philippines China Singapore 12 5.5 3-4 3 27 30 Papua 6 31 18 it Melanesia Micronesia 0-29 0-9 “Key to references: 1. Marie bin coed SL en + : + LM 28.29 30 13. Plato et al. (1964); N 17. 16. Szeinberg and Sheba (1960); 15. Bogair (1951); 14. Szeinberg et al. (1958b); Efrati (1952); 18. Taleb er al. (1964); 19. Ragab er al. (1966); 20. Say et al. (1965); 21. Amin-Zaki ef al. (1972); 22. Shaker et al. (1966); 23. Gelpi (1965); 24, Walker | È { and Bowman (1959); 32 33 9, Crosby (1956); Ì TABLE I (Continued) 11. Zannos-Mariolea and Kattamis (1961); (1964); (1961); 3. Motulsky + al. (1969); i | à Gilerati 11660 N | a Klon etal. (1961); aan rn dal (1961); || 35.26 10. Luisada (1941); 12. Kattamis et al. (1969); 25. Hedayat er al. (1971); 26. Belsey (1973); 27. Chan ef al. | 28. Du(1952); 29. Vella(1959); 30. Wong(1972); 31. Ryan and Parsons 32. Kidson and Gorman (1962); 33. Kidson and Gajdusek (1962). | i è Plus (+) sign indicates that outbreaks of favism have been reported in the respective populations; blank space indicates that, to our knowledge, no such occurrences have been reported.

By the use of ingenious cytochemical techniques capable of detecting G6PD a deficiency in single cells, it could be shown that erythrocytes of females with of mixture a e constitut cy deficien heterozygous trait of intermediate enzyme normal and G6PD-deficient cells (Beutler er al., 1962; Sansone et al., 1963). This so-called cellular mosaicism bears out the prediction of Lyon's ‘‘X chromosome inactivation’’ theory (Lyon, 1961). According to this concept, a random tic congenital hemolytic anemia,’’ occurring spontaneously in the absence of any extraneous challenging agent (Yoshida, 1970). The most common G6PD-deficient mutants are as follows: the Mediterranean type, which is prevalent among Sephardic Jews, Italians, and Greeks; the A type a among Negroes; the Canton type found primarily among southern Kaann other Oriental populations; and the Debrousse type most frequent of maternal origin) takes place in each individual somatic cell early in the course It progeny. the in pattern ion inactivat the morphogenesis, with a perpetuation of favic or uced drug-ind to ility susceptib of follows, therefore, that the degree hemolysis in heterozygous females will be critically dependent on the relative proportions of the G6PD-deficient and normal erythrocytes in their blood. Three variants of human G6PD that were isolated in molecularly homogeneous form and subjected to ‘‘fingerprinting’’ were found to differ from one another in a single amino acid residue. Thus, an asparagine residue present in variant B is 272 loss of functional activity of one of the X chromosome pair (of either paternal or 273 replaced by aspartic acid in variant A (Yoshida, 1967a). Similarly, G6PD Hektoen was found to differ from the B variant by a substitution of histidine for tyrosine (Yoshida, 1970). During the last decade, G6PD has been the subject of intensive enzymological investigations. The normal enzyme was found to exist in several oligomeric forms (composed of two, four, or six identical subunits), the degree of aggregation depending on a variety of physicochemical factors, e.g. ionic strength, pH, and protein concentration of the enzyme solution (Yoshida, 1966, 1967b; Yoshida and Hoagland, 1970). The catalytically active molecular species that appears to be predominantly in the dimeric form is stabilized by its tight associathe tion with NADP or NADPH, whereas in the absence of the coenzymes 1973). , (Yoshida s monomer inactive the into e apoenzyme tends to dissociat Starch gel electrophoresis of normal red cell hemolyzates revealed the existence of two major molecular variants of G6PD, characterized as a fast migrating form B is band A and a slow band B (Boyer et al., 1962). The more common found in Caucasian subjects and in American Negroes, whereas type A occurs among Negroes only (Boyer et al., 1962; Kirkman and Hendrickson, 1963). In subsequent studies, about 80 additional genetic variants of G6PD were differentiated by a combination of various enzymological criteria, such as catalytic rate, electrophoretic mobility, substrate specificity, and particularly the ability to utilize 2-deoxy-D-glucose, Km values for glucose 6-phosphate and NADP, pH optimum, and heat stability (Kirkman er al., 1964; Yoshida et al., 1971; Beutler, 1971). In nearly 40 variants that are not associated with any clinical symptoms, the catalytic activity is within the normal range and in one instance (G6PD Hektoen) is even severalfold higher than the normal average level. About 20 variants anemia exhibit severe red cell enzyme deficiency, which manifests as hemolytic comprisgroup, Another beans. fava or drugs by stress in response to oxidative ing about 20 G6PD-deficient mutants, is associated with ‘‘chronic nonspherocy- By applying a density gradient centrifugation procedure to separating red blood cells into different age groups, it was found that the G6PD activity in a 5% fraction of youngest cells from a blood sample of the deficient A- variant was practically identical to that of normal erythrocytes (Yoshida eral., 1967). It was concluded that the G6PD A” enzyme has normal initial specific activity, but the rate of its inactivation and eventual loss from the cell is considerably enhanced This conclusion was further corroborated by immunological methods (Marks and Gross, 1959; Piomelli et al., 1968; Yoshida et al, 1968). On the other hand, by the use of similar procedures, the Mediterranean mutation was found to involve both a decrease in the number of molecules and a diminished specific activity of the enzyme (Yoshida et al., 1968). G. Correlation between the Degree of G6PD Deficiency and the Severity of Clinical Symptomatology The severity of clinical manifestations among the various G6PD mutants does not correlate well with the differences in the extent of G6PD deficiency, as measured in vitro under the conventional assay conditions. Yoshida and Min ( 1973) suggested that the apparent inconsistencies could be resolved by taking into account the regulatory mechanisms governing the G6PD activity under the physiological conditions prevailing within the red blood cell. It was found that the activity of the oxidative pentose phosphate shunt in the red cell is strongly suppressed, representing only about 0.1-0.2% of the maximal potential catalytic rate of G6PD, as determined in cell-free hemolyzates. The low intracellular activity of the pentose shunt is attributable to the strong inhibitory effect of NADPH and ATP on the activity of G6PD, which constitutes the rate-limiting step in the overall oxidative pentose phosphate pathway. The various G6PDdeficient mutants differ in their affinities for NADP (K,,), as well as in the degree F. Molecular Characteristics of Normal and Mutant G6PD

J. MAGER, M. CHEVION, AND G. GLASER H (K;). Thus, the sponof their susceptibility to inhibition by ATP and NADP Tripler, or Alhambra) are taneously hemolytic variants (e.g., Manchester, the conditions existing in strongly inhibited by ATP and NADPH, so that under 1:50 and 1:5 mM ATP) to the human erythrocyte (NADP/NADPH ratio close potential G6PD activity their gh these variants are virtually nonfunctional, althou nonhemolytic mutants (e.g., is about 20% of the normal. In contrast, the H and ATP, and, thereNADP by Mediterranean or A”) are resistant to inhibition catalytic activity varies fore, under simulated physiological conditions, their major variants (A and normal the of between 30 and 50% of the average activity B). H. Differences between Caucasian and Negro Types Deficiency of G6PD heterogeneity of the The considerable genetic polymorphism and biochemical least partly account at may enzymatic defect brought to light by the above studies clinical manifestaand ical for the marked differences and variations in the biolog differences are The . groups tions of the inborn error observed in different ethnic Caucasians and s Negroe can particularly striking between the affected Ameri (Italians, Greeks, Sephardic Jews, etc.). Thus, in 275 lation with the frequency of exposure to Vicia faba or the degree of enzyme deficiency. A study conducted by these authors on the incidence of favism in three different areas of Greece showed that overt episodes of favism do not occur at random in G6PD-deficient subjects, but there is a tendency for regional and familial aggregation of cases. Thus, in the area of Karditsa with a frequency of G6PD deficiency of about 27%, favism is rare, whereas on Corfu Island with an approximately 5% incidence of this enzyme deficiency, favism is of relatively frequent occurrence. In both areas, the consumption of fava beans is very common. The authors concluded that the family data they collected are consistent with ‘‘the hypothesis of Mendelian segregation of an autosomal gene which in the heterozygous state enhances the susceptibility to favism of G6PD-deficient individuals. *” The functional role of the hypothetic gene remains to be defined Its influence on the susceptibility of the red blood cells to hemolysis or aftematively its modifying effect on the absorption, detoxication, or excretion of the causative agent of favism are among the various possibilities to be considered (Stamatoyannopoulos et al., 1966; see also Tarlov et al., 1962). A preferential association between favism and certain acid phosphatase phenotypes (types A and C) has been described (Bottini et al., 1971). However, the pathogeneti significance of this correlation is not clear. ns o G6PD-deficient Negro males, % of the normal mean, the range of the erythrocytic G6PD activity is about 10-21 of the normal level 0-6% only is y whereas in Caucasian males the enzyme activit (Marks and Gross, 1959; Marks and Banks, 1965). This disparity in the extent of relatively milder clinical the enzyme deficiency seems to be correlated with the in Negroes (see Beutler, ed observ course of the drug-induced hemolysis usually activity of the young G6PD the 1972; Pannaciulli et al., 1965). Furthermore, erably elevated in consid is crisis) red cell population (following a hemolytic The enzyme defi1965). Banks, and Negroes but not in Caucasians (see Marks of affected tissues nt differe the in ciency appears to be more widely distributed Chan et al., 1965; Marks et Caucasians than in Negroes (Brunetti ef al., 1960; ians is usually repreal., 1959). The vestigial G6PD activity in deficient Caucas enzyme shows invarthe s Negroe nt deficie sented by the variant B, whereas in 1962). al., iably A-type characteristics (Boyer et Than I. Possible Role of Genetic Determinants Other Deficiency in the Pathogenesis of Favism G6PD ated by StamatoyanThe existence of an extracorpuscular factor was postul nce of favism in incide low ely relativ the n nopoulos et al. (1966) to explai s, as well as its Negroe can Ameri in e absenc G6PD-deficient subjects, its total shows no correwhich ence, occurr of mode e bizarre and apparently unpredictabl J. Methods of Detection of G6PD Deficiency 1. Glutathione Stability Test The test designed by Beutler (1957) is based on his observation that the GSH content of G6PD-deficient but not of normal erythrocytes declines markedly in the course of aerobic incubation with acetylphenylhydrazine and glucose under standardized conditions. The recommended method for glutathione determination is the procedure of Ellman (1959) as adapted by Beutler ef al. (1963). 2. Spectrophotometric Determination of G6PD on assay is based on the measurement of the rate of increase of absorbance at nm due to formation of NADPH in a reaction system consisting of glucose ’ > > 3. Methemoglobin Reduction Test 4 The assay introduced by Brewer et al. (1962) is based on the observation of awson et al. (1958) that the rate of methemoglobin reduction by G6PD-

deficient erythrocytes in the presence of methylene blue is considerably slower e blue), than normal. A modification using Nile blue sulfate (instead of methylen detecthe for designed was cells, between on a dye that does not permit interacti (Beutler females gous heterozy in y deficienc tion of intermediary states of G6PD and Baluda, 1963). 4. Indicator-Linked Screening Methods y by In these tests, the G6PD-linked reduction of NADP is measured indirectl following the rate of reduction of a suitable dye serving as an artificial terminal hydrogen acceptor. The original procedure of Motulsky and Campbell-Kraut was (1961), in which the decolorization time of brilliant cresyl blue is determined, modified by the use of 3-(4,5-dimethylthiazolyl-1,2)-2,5-diphenyltetrazolium a purple inbromide (MTT), which is reduced by the G6PD-linked system to techsoluble formazan derivative. The latter reagent can be used in a spot-test nique (Fairbanks and Beutler, 1962). 5. Fluorescent Screening Methods This screening procedure, introduced by Beutler and Mitchell (1968), takes advantage of the fact that NADPH (generated by G6PD) fluoresces when illuminated with ultraviolet light, whereas NADP does not. A small volume of blood is hemolyzed with saponin and mixed with a buffered solution of suitable amounts of glucose 6-phosphate, NADP, and GSSG. Following incubation for 5-10 min, the mixture is spotted on filter paper and examined under an ultraviolet lamp for fluorescence. An automated, quantitative version of this method adapted to use with the Technicom AutoAnalyzer instrument has been described (Dickson et al., 1973). 6. Ascorbate-Cyanide Test This test, devised by Jacob and Jandl (1966), depends on the ability of ascorof bate to undergo oxidation in the presence of oxyhemoglobin with formation detoxbe can H,O, generated the cyanide, by inhibited is H,O,. When catalase the ified only by GSH peroxidase. Any enzymatic deficiency interfering with , reductase GSH GSH, (G6PD, system e peroxidas GSH the proper functioning of and GSH peroxidase) causes H,O, to accumulate. The hydrogen peroxide reacts with hemoglobin, producing hemochromes with resultant brown discoloration detectable by visual inspection. 277 ll. THE SELECTIVE TOXICITY OF FAVA BEANS: SEARCH FOR THE CAUSATIVE AGENT OF FAVISM A. Effect of Ingestion of Fava Beans on the Life Span of G6PD-Deficient Erythrocytes Efforts to obtain direct experimental evidence for the selective toxicity of broad beans for G6PD-deficient cells are greatly handicapped by the lack of a susceptible laboratory animal and the potential health risk inherent in the induction of a hemolytic disorder in sensitive volunteers. To circumvent this difficulty, some authors have adopted the experimental stratagem of Dern et al. (1954), consisting of the transfusion of 5!Cr-labeled G6PD-deficient erythrocytes into normal compatible recipients and the determination of the effect of orally administered broad beans: on the survival of the tagged cells. These studies have yielded conflicting results in the hands of different investigators. Thus, both Greenberg and Wong (1961) and Davies (1962) failed to detect any significant decrease in the life span of the transfused erythrocytes following ingestion of broad beans by healthy recipients. The negative outcome of these trials, however, is open to criticism and may be ascribed perhaps to the inadequacy of the experimental conditions, such as insufficient size of the challenging dose of the fava beans or loss of their noxious activity due to cooking or prolonged storage. On the other hand, Panizon and Vullo (1962) concluded from their welldocumented experiments that fresh fava beans or juice prepared from them induced in the majority of cases a definite shortening of the survival of the transfused G6PD-deficient erythrocytes. According to the authors’ estimate, however, the effective dose of the fava beans was surprisingly high, i.e., about 50,000 times larger, on a weight basis, than the minimal hemolytic dose of primaquine (Panizon and Vullo, 1962). Furthermore, the hemolytic effect of the fava beans was much more pronounced when fava-sensitive individuals (in the early stage of recovery from a favic crisis) rather than normal volunteers served as recipients. These results, therefore, seem to indicate that G6PD-deficiency and ingestion of fava beans are not sufficient by themselves to bring about a full-blown hemolytic crisis and point to the adjuvant role of an extracorpuscular factor in the pathogenesis of this disease (Panizon and Vullo, 1961, 1962; Panizon, 1967). B. Studies with Crude Fava Bean Extracts Attempts to ascertain unequivocally the presence of a noxious factor in fava beans with a specific effect on drug-sensitive erythrocytes have thus far met with a moderate degree of success.

Mela and Perona (1959) noted a decrease of GSH in G6PD-deficient red blood with a cells incubated in the presence of broad bean juice prepared by squeezing exhydraulic press. Walker and Bowman (1960) reported that crude aqueous (Bowman tracts of fava beans, as well as saline extracts of fava pollen and pistils GSH level the in drop nt significa and rapid a caused 1961), Walker, and tes. of sensitive erythrocytes but had little or no effect on normal erythrocy No min. The activity of the broad bean extracts was destroyed by boiling for 90 seeds us effect was detected with extracts prepared from peas or other legumino tested. Results similar to those described above, though less clear-cut, were also obtained by Contu ef al. (1961). Furthermore, Panizon and Zacchello (1965) found that fava bean juice or whole bean homogenates induced a fall of GSH and a definite impairment of the survival of °'Cr-labeled G6PD-deficient red blood cells transfused into normal recipients, whereas normal erythrocytes remained personal unaffected. Panizon and his collaborators (Panizon, 1967; F. Panizon, in lipid communication, 1968) isolated from fava beans a material soluble solvents that induced a very marked decline of the intraerythrocytic GSH accomto panied by formation of methemoglobin and Heinz bodies; in contrast The in. primaquine, this material also acted on GSH in the absence of hemoglob of authors concluded that their data are incompatible with an allergic patho genesis favism, a view still held by certain investigators (Carcassi, 1958; Kantor et al., s 1962). They also suggested that fava beans and primaquine give rise to hemolysi ashis and through an essentially similar mechanism, More recently, Bottini sociates (Bottini et al., 1970; Bottini, 1973) described the separation of crude in fava bean extracts into two distinct fractions, both capable of oxidizing GSH pure solution and in G6PD-deficient erythrocytes. Various serum abnormalities were observed in patients during an acute favic episode, such as atypical antibodies against their own erythrocytes (Marcolongo n, et al., 1950) or antibodies against fava bean extracts (Kantor and Arbesma 1959; Kantor et al., 1962). Some investigators described the occurrence in saline extracts of broad beans of a heat-labile substance causing agglutination of both normal and G6PD-deficient erythrocytes (Roth and Frumin, 1960; Greenberg and Wong, 1961) and counteracted by a normal serum factor residing in the IgA fraction of the y-globulins (Creger and Gifford, 1952). Frumin and his coworkers reported that the factor neutralizing the fava bean hemagglutinin was absent in the serum of a number of favic patients tested both in the acute hemolytic stage and during the remission (Perera and Frumin, 1965; Nathan et al., 1974). These findings are at variance with the observation of Greenberg and Wong (1961) that both favic and normal sera exhibited the ability to inhibit hemagglutination by fava bean extracts. The paucity of the data does not permit a critical assessment of their significance and their possible relevance to the pathogenesis of favism. 279 C. Fractionation of Fava Bean Extracts In a study carried out in our laboratory, the quest for the causative agent of favism in broad bean extracts was guided by its presumed discriminatory capacity for oxidizing GSH in G6PD-deficient but not in normal erythrocytes, when incubated in vitro in the presence of glucose. In fact, several fractions conforming to this criterion could be isolated from aqueous broad bean extracts by use of ion-exchange chromatography. Some of the purified fractions were sparingly soluble in water and exhibited in neutral solution a rapid loss of GSH-oxidizing activity, concomitant with a change in their spectral characteristics. The structural instability of these substances appeared to be inherent in their tendency to undergo spontaneous oxidation in air, since the deterioration could be largely prevented by storage under nitrogen (S. Bien, M. Noom, G. Glaser, A. Roigin and J . Mager, unpublished results). The properties of the active fractions were reminiscent of those described for some pyrimidine derivatives known to occur in fava beans in the form of aglycones of the B-glycosides termed ‘‘vicine’’ and ‘‘convicine.’’ Our subsequent work, therefore, was concerned primarily with exploring the possible role of these aglycones in the causation of favism. D. Structure and Properties of Vicine, Convicine, and Their Aglycones Vicine was first isolated by Ritthausen and Kreusler (1870) from Vicia sativa seeds by a method involving extraction with dilute H,SO, and precipitation with HgSO,; the final yield of the crystalline material was about 0.35% (Ritthausen 1876). Vicine was subsequently found in other species of Vicia including Vicia faba (Winterstein and Somló, 1933), beet juice, and peas (Schulze, 1891; Bendich and Clements, 1953). The glycosidic nature of this compound was recognized by Ritthausen (1896), who also succeeded in isolating the aglycone divicine (Ritthausen, 1899a,b). The pyrimidine nucleoside structure was assigned to vicine by Johnson (1914) and confirmed by Levene (1914). The correct formulation, however, of vicine as 2,6-diamino-4,5-dihydroxypyrimidine 5-(8-D-glucoa © (D was arrived at only several decades later by Bendich and Clements Convicine was also discovered by Ritthausen (1881) in Vicia sativa. It was identified by Johnson (1914) as a B-glycoside of isouramil, and the correct position of the glycosidic bond was established by Bendich and Clements (1953). The formulation of convicine as 2,4,5-trihydroxy-6-aminopyrimidine 5-(B-b-glucopyranoside) (II) was confirmed by unambiguous evidence (Bien et al., 1968). The aglycones divicine (IID and isouramil (IV) can be obtained from the respective glycosides (vicine and convicine) by mild acid hydrolysis or by enzymatic splitting with 8-glucosidase (emulsin). In recent years, several synthetic

iT OH N; 2 1 H,N“ 6 "N Vicine Convicine (1) Mm) H‚N NZ OH N NH, HO N NH, (III) (IV) C—OH —C— OH —C—NH, A B Il reducing properties, spectral characteristics, and molecular instability, are strikOH Isouramil | pod Bendich and Clements (1953) pointed out in their monumental study that some of the distinctive features of divicine and its cogeners, namely, their powerful NZ Divicine — C=0 in nonspecific ‘‘end absorption’’ occurring in the course of the above-mentioned oxidative decomposition of these compounds are suggestive of the rupture of the pyrimidine ring structure. OH OH = 281 ingly similar to those of ascorbic acid. Furthermore, these authors inferred from a ed (Davoll and procedures for preparing these compounds have also been develop ef al, 1964; Bien et Laney, 1956; McOmie and Chesterfield, 1956; Chesterfield al., 1968; Ikeda et al., 1973). with the Vicine and convincine, as well as the corresponding aglycones, react d by produce that to similar color Folin-Ciocalteu phenol reagent, yielding a blue (1974) Read and Higazi by d tyrosine or tryptophan. This property was employe and as a basis for developing a quantitative assay of vicine in plant material blood. In our hands, however, the method failed to fulfill the authors’ claim for of the its specificity and, therefore, did not lend itself to direct determination reduce l isourami and divicine Both pyrimidine glycosides in natural materials. molybvigorously alkaline solutions of 2,6-dichlorophenolindophenol, phospho amwith reaction color blue date, or phosphotungstate and elicit an intense l hydroxy enolic an of e presenc the of ve moniacal ferric chloride solution, indicati of e presenc the in unstable highly are group in the molecule. The aglycones pH and oxygen; the rate of their oxidative breakdown is most rapid at alkaline es of half-liv the ure, temperat room At falls off with decreasing pH values. The min. 30-40 of order the of are s divicine and isouramil in neutral solution breakdown of the pyrimidine aglycones is accelerated by traces of copper (Cu?*) destroyed and other heavy metals; both compounds are almost instantaneously by boiling (Bendich and Clements, 1953; M. Chevion et al., results to be published). characThe closely similar ultraviolet spectra of divicine and isouramil are to oxygen to e exposur upon shifting form), d terized by a peak at 280 nm (reduce al., ef (Razin form) d (oxidize nm 255 at m a less prominent absorption maximu increase 1968). The disappearance of the characteristic peak and the concomitant study of different substitutions that the common structural denominator underlying these properties is a carbonyl-conjugated enediol (A) or aminoenol (B) system. Consequently, all the characteristic properties of the aglycones are abolished by substitution of the hydroxyl group at C-5, such as that represented by the glycosidic linkage present in vicine and convicine. In fact, these glycosides show none of the reducing properties of their aglycones, are remarkably heat stable in solution, and their ultraviolet spectra differ significantly from those of their constituent pyrimidines (Bendich and Clements, 1953; Bien et al. 1968; S. Bien, M. Noam, G. Glaser, A. Razin, and J. Mager, anpublished results). E. Effects of Divicine and Isouramil on Red Cell Metabolism in Vitro: Synergistic Interaction of Isouramil and Ascorbic Acid Incubation of human red cell suspensions in phosphate-buffered isotonic saline (pH 7.4), supplemented with isouramil or divicine (to be referred to as aglycones), resulted in a rapid fall of their GSH level, followed by a slower decline of their ATP content (Mager et al., 1965). Addition of glucose prevented the injurious action of the aglycones on normal erythrocytes but had no protective influence on G6PD-deficient cells. This behavior of the aglycones contrasted with the virtual inertness displayed in the same system by convicine and vicine in which the readily oxidizable enolic hydroxyl group at C-5 of the pyrimidine motety is blocked by the B-glycosidic bond. The activity of the aglycones, as gauged by their effects on the GSH and ATP contents of the erythrocytes na roughly 20-30 times higher than that of acetylphenylhydrazine (APH). m contrast, the structurally related pyrimidine, dialuric acid, was relatively ineffective in the test system used, presumably due to its pronounced tendency to interact with GSH by producing an addition compound with a characteristic absorption peak at 305 nm (Patterson ef al., 1949). In contradistinction to APH and primaquine, which require the presence of

J. MAGER, M. CHEVION, AND G. GLASER hemoglobin for catalyzing GSH oxidation (Beutler et al., 1957), the fava n pyrimidine aglycones were able to oxidize GSH in pure solution. The oxidatio higher or tenfold a at even on completi to ed of GSH by these compounds proceed in air molar ratio of GSH to pyrimidine; it was considerably enhanced by shaking c hiometri nonstoic The nitrogen. of ere atmosph and completely suppressed in an the by GSH of n oxidatio the of nce nature, as well as the oxygen depende uction aglycones, suggested that this reaction is mediated by a catalytic oxidored system redox acid orbic GSH-asc own well-kn the mechanism, similar to that of the observation (Borsook et al., 1937). This analogy is further emphasized by that sodium ascorbate at relatively high concentrations (5-10 nM) was capable of producing a marked decrease (50-80%) in the GSH content of G6PD-deficient erythrocytes incubated in phosphate-buffered saline solution (pH 7.4) for 3 hr at Moreover, a 37°C (see also Waller and Benöhr, 1973: Prins and Loos, 1969. ately mixture of 0.05 mM isouramil and 0.5 mM ascorbate caused an approxim on conditions, 70% decline of the intracellular GSH level under similar incubati ve in whereas the same amounts of each compound alone were totally ineffecti human or this system. Essentially similar results were obtained with normal of absence the in out carried was on incubati the that provided rabbit erythrocytes, ascorand l isourami of affect d combine the of nature itive supraadd glucose. The of bate on erythrocyte GSH was shown to represent a net outcome of a series rapidly the with on interacti direct its by GSH of n oxidatio namely, reactions, autoxidizable isouramil and the concomitantly generated H,O,, as well as indiby rectly by a cyclic system involving oxidation of ascorbate to dehydroascorbate GSH by e ascorbat to scorbate dehydroa of n reductio renewed the isouramil and tionship (Razin et al., 1968). It is not unlikely that such a synergistic interrela may play a crucial role in the pathogenesis of favism. d) In a subsequent series of experiments (G. Glaser er al. results to be publishe designed to test the effect of isouramil on red cell survival, glucose-deprived ed rabbit erythrocytes were labeled with °'Cr and incubated in phosphate-buffer of saline at 37°C for different periods of time in the presence of various amounts isouramil. The cells were then reinfused into the donor rabbits, and their °'Cr half-life was determined. It was found that incubation of the erythrocytes with 2 mM isouramil for 75 min resulted in a reduction of their Cr half-life to less than l 5% of the normal value (9 hr and 9 days, respectively). Lower levels of isourami or shorter incubation periods gave rise to correspondingly less pronounced iml pairment of the red cell survival. Furthermore, here again the effect of isourami was found to be synergistically enhanced by addition of ascorbate. 283 of vicine (0.2 gm/kg body weight), which was isolated from Vicia faba by a modified procedure of Levene (1914; Lin and Ling 1962a). These authors found also that vicine exerts some minor inhibitory effects on the activities of glucose-6-phosphate and 6-phosphogluconate dehydrogenases in human red cell hemolysates (Lin and Ling, 1962c). However, the significance of these effects in the pathogenesis of favism is rather doubtful. In another paper, Lin (1963) described some chemical interactions between divicine and sulfhydryl compounds. With an excess of GSH relative to divicine a maximal absorption appeared at 305 nm, whereas with a mixture of Meine and cysteine two peaks, at 285 and 245 nm, were observed. In both instances, the amounts of sulfhydryl compounds that disappeared were severalfold higher où a molar basis, than the quantity of divicine added. In ‘our opinion, the complex absorbing at 305 nm plays no essential part in mediating the catalytic oxidation of GSH by divicine, but rather represents a side reaction analogous to that observed with alloxan (Patterson et al., 1949). Furthermore, according to our data, the spectrum of the pyrimidine-cysteine mixture appears to be determined by the | . reducin molecular structure of divicine per se stabilized by the of cysteine action of action g y (Razini et al., 1968). F. Possible Etiological Role of Divicine and Isouramil in Favism The overall pattern of metabolic disturbances resulting from incubation of G6PD-deficient red cells with the aglycones of vicine and convicine is essentially identical to that elicited by treatment with acetylphenylhydrazine (Beutler et al. 1957; Mager et al., 1965). The powerful capacity for oxidizing GSH exhibited by the pyrimidine aglycones in vitro as well as the observed deleterious effect of isouramil on red cell survival are consistent with a possible causative role of these substances in precipitating the favic crises. The free aglycones may arise from the parent fava glycosides either in the beans or in the digestive tract through the hydrolytic action of B-glucosidase. The conceivable vicissitudes in the availability of the requisite conditions for enzymatic release of the aglycones from the glycosides, as well as the particular lability of these compounds, might account for the puzzling irregularity that characterizes the occurrence of favism in susceptible individuals, irrespective of the degree and frequency of their exposure to the noxious agent (see Luisada, 1941). It may be pertinent to mention in this connection that divicine has been in the past named as the causative agent of neurolathyrism (Anderson et al., 1925) earlier Shortly after the publication of our study (Mager et al., 1965), the because of its parenteral toxicity to experimental animals (Kleiner, 1912). This view, however, was discredited by the finding that oral administration of divicine observed a Abstracts in 1966. The Formosan workers (Lin and Ling, 1962b) Harper, 1963) produced no adverse effects other than growth retardation. As Pointed out by Liener (1966), it was probably because of this lack of specific to investigations of Lin and Ling (1962a,b,c) on the possible relation of vicine l Chemica favism, came to our attention through summaries appearing in transient hemoglobinuria occurring in puppies 3 hr after the oral administration at a level as high as 1% of the diet to rats (Lee, 1950) and chicks (Arscott and

toxicity of divicine by oral route that its possible significance as the causative principle of favism has escaped the attention of the earlier investigators in this field. uct of dopa, rather than dopa itself is the active factor responsible for the fava bean-induced hemolysis. This claim is based on the observation that a mixture of dopa (0.15 mM) and tyrosinase (presumed to generate dopaquinone) caused a rather inconspicuous decrease of the GSH content in G6PD-deficient but not in normal erythrocytes. However, the postulated role of dopa in the pathogenesis of favism was not borne out by the outcome of an experiment in vivo indicating that the survival of *'!Cr-labeled G6PD-deficient erythrocytes transfused into a normal individual was not impaired by repeated intravenous administration of dopa to 284 G. Effect of 3,4-Dihydroxyphenylalanine (Dopa) on Erythrocyte GSH and Critical Evaluation of Its Postulated Role in the Etiology of Favism Kosower and Kosower (1967) put forward the hypothesis that 3,4-Ldihydroxyphenylalanine (dopa) may be one of the active principles responsible for the ability of fava beans to induce hemolysis in G6PD-deficient individuals. This substance, known to be a moderately strong reducing agent, is present in broad beans in substantial amounts [about 0.25% of the fava pods (Guggenheim, 1913)], mainly in the free state and partly in the form of its B-glycosidic derivative (Pridham and Saltmarsh, 1963; Andrews and Pridham, 1965). Kosower and Kosower (1967) found that significant losses of GSH occurred in G6PD-deficient erythrocytes, when incubated at 37°C for 3 hr in a glucose-containing medium supplemented with dopa in amounts ranging from 0.75 to 3 wmoles/ml. In contrast, oxidation of GSH by dopa in normal red blood cells was demonstrable only in the absence of glucose. Careful scrutiny of the data of Kosower and Kosower reveals that the amounts of GSH that disappeared (were oxidized) were related to the amounts of dopa added by a roughly 1:10 molar ratio. Thus, contrary to the authors’ claim, these results do not seem to support the notion of a nonstoichiometric (catalytic) oxidation of GSH by dopa. The latter conclusion is also in line with our observation that no appreciable oxidation of GSH took place when a mixture of GSH (2 mM) and dopa (4 mM) in 0.01 M phosphate buffer (pH 7.4) was incubated for 30 min at 37°C with continuous shaking in air. Moreover, comparative experiments showed that dopa at a concentration as high as 10 mM failed to affect the GSH level in normal washed human erythrocytes incubated for 3 hr at 37°C in the absence of added glucose, whereas 1 mM isouramil caused almost complete dissappearance of the intracellular GSH under the same conditions. On the other hand, combined addition of 1 mM dopa and 0.2 mM isouramil resulted in nearly 80% destruction of the erythrocytic GSH, whereas each compound alone was without perceptible effect (Razin et al., 1968). Similar results were obtained in experiments performed with glucose-starved rabbit erythrocytes. In addition, however, it was found that, although dopa potentiated the oxidant action of isouramil on the intracellular GSH, it failed to enhance the effect of isouramil in shortening the survival of *'Cr-tagged erythrocytes treated in vitro and reinfused into the donor rabbit (G. Glaser et al., results to be published). More recently, the hypothesis of Kosower and Kosower (1967 was endorsed by Beutler (1970) and modified to suggest that dopaquinone, the oxidation prod- 285 the recipient (Gaetani et al., 1970). Furthermore, as pointed out by Beutler himself, since L-dopa is being used extensively in rather large doses for treating Parkinson’s disease, some of the Parkinsonian patients (with coexistent G6PDdeficiency) in favism-prone areas would be at risk of developing hemolytic crises in the course of therapy. To our knowledge, however, so far no single case has been reported in the literature to substantiate this expectation. IH. THE MECHANISM OF THE BIOCHEMICAL LESION UNDERLYING RED CELL DESTRUCTION IN DRUG-INDUCED HEMOLYSIS AND FAVISM The hemolytic effect of the noxious drugs on G6PD-deficient erythrocytes appears to be attributable to their ability to function as reversible redox systems mediating the oxidation of the intracellular GSH (Emerson et al., 1949; Beutler et al., 1957). This property is shared also by divicine and isouramil, the pyrimidine aglycones of the fava bean glycosides vicine and convicine. In the normal red blood cell the oxidant effect of the drug is readily overcome by the coordinate action of the NADPH-generating pentose phosphate pathway and the NADPH-linked GSSG-reductase according to the following reaction scheme: [Eqs. (1)-(3)]: dehydrogenase Glucose 6-phosphate + NADP* ———————————_> 6-phosphogluconate + NADPH + H+ dehydrogenas 6-Phosphogluconate + NADPt à () (2) ribulose 5-phosphate + NADPH + H+ + CO, GSSG + NADPH + H* reductase GSH idas: 2GSH + NADP+ 2GSH + HO, MST. — GSSG + 2H,0 (3) Under physiological conditions, the vestigial G6PD activity and perhaps also the limited capacity of GSSG reductase to use NAD as an alternate hydrogen

J. MAGER, M. CHEVION, AND G. GLASER donor (Francoeur and Denstedt, 1954) enable the enzyme-deficient erythrocyte to maintain an adequate level of GSH compatible with a nearly normal or moderately reduced survival (Brewer ef al., 1961). This precarious metabolic equilibrium, however, breaks down under the stress conditions imposed by the oxidant compound. The resultant irreversible oxidation of GSH and the attendant catabolism of GSSG (Beutler, 1957) seem to constitute the major metabolic lesion leading to the eventual destruction of the enzyme-deficient erythrocyte. The validity of this concept, implying a vital role of GSH in preserving the structural integrity of the red blood cells, is strongly supported by the finding that the virtual absence of GSH in the blood cells of individuals affected with an inborn defect of its biosynthesis predisposes them to drug-induced hemolysis and favism (Oort et al., 1961; Waller and Gerok, 1964; Boivin and Galand, 1965; Prins et al., 1966; Minnich et al., 1971). Similarly, congenital GSSG-reductase deficiency likewise manifests itself by drug sensitivity (Loehr and Waller, 1962; Waller et al., 1965, 1969). A major manifestation of the oxidant action of the drugs both in vivo and in vitro is the formation of methemoglobin and the concomitant appearance of Heinz bodies, which, according to Allen and Jandl (1961), represent a product of hemoglobin denaturation resulting from oxidation of its SH groups with concurrent formation of a mixed glutathione disulfide and loss of the heme group (see also Srivastava and Beutler, 1970; Bunn and Jandl, 1966; Jacob and Winterhalter, 1970; Jacob, 1970; Rachmilewitz ef al., 1969; Nagel and Ranney, 1973). Cohen and Hochstein (1961, 1963, 1964) indicated that the oxidant drug or its active metabolite interacts with oxyhemoglobin, producing hydrogen peroxide. The relatively low but potentially harmful levels of peroxide cannot be efficiently destroyed by catalase and are normally eliminated through the action of GSH peroxidase (Mills, 1957, 1959, 1960; Mills and Randall, 1958), which catalyzes the following reaction [Eq. (4)]: 1970; Boivin et al., 1969, 1970). Furthermore, the recent discovery that GSHperoxidase contains selenium as an integral and catalytically essential component of its molecule (Rotruck et al., 1973; Flohé et al., 1973) has led to the understanding of the biochemical mechanism underlying the protective effect of dietary selenium against hydrogen peroxide-induced hemolysis (Rotruck et al., 1972). Jacob and Jandl (1962a,b), in studying the effects of SH-binding compounds (p-hydroxymercuribenzoate, N-ethylmaleimide) on erythrocytes, emphasized the essentiality of the surface sulfhydryl groups for the structural intactness and normal survival of these cells. It should be pointed out, however, that contrary to the typical thiol reagents used in the above studies, primaquine and related drugs (Panizon and Zacchello, 1966; Beutler, 1966), as well as the fava bean pyrimidine aglycones (Mager et al., 1965), do not induce an overt lysis in vitro but appear to exert their deleterious effect in vivo by rendering the red cells vulnerable to destruction by the reticuloendothelial system in the liver and spleen (Rifkind, 1965, 1966; Beutler, 1971). Consequently, it is not clear to what extent the conclusions drawn from the model experiments of Jandl and his associates are applicable to drug-induced hemolysis and favism. Some investigators (Kosower ef al., 1969; Flohé er al., 1971) suggested that in certain instances free radicals, rather than H,O,, generated in the course of the metabolism of the noxious agent, may interact with GSH and protein thiols with the resultant formation of GSSG and protein S—S linkages. Other studies have been concerned with the possible derangement of the energy-yielding metabolism as part of the mechanism underlying the druginduced red cell hemolysis. It was observed by several authors that aerobic incubation of erythrocyte suspensions in the presence of primaquine or acetylphenylhydrazine resulted in a pronounced inhibition of glycolysis (Loehr and Waller, 1961; Kosower et al., 1964) and a progressive decrease in the ATP level of the cells (Mohler and Williams, 1961; Loehr and Waller, 1961; Mager et al., GSH-peroxidase 2GSH + H,O, 287 GSSG + 2H,O (4) The sustained operation of this system is ensured by the concomitant regeneration of GSH, mediated by the NADPH-linked GSSG-reductase. Thus, the integrated pathway consisting of the oxidative pentose phosphate shunt (as a source of NADPH supply), GSSG-reductase, and GSH-peroxidase serves to detoxify the hydrogen peroxide, so as to obviate its deleterious effects on the red cell membrane and hemoglobin (see also Cohen, 1966; Flohé and Brand, 1969). This concept, assigning a vital function to GSH-peroxidase in protecting the cell from the oxidative insult by the peroxide-forming drugs, has gained additional support from the recognition of a hereditary deficiency of this enzyme and its causative role in certain cases of spontaneous or drug-induced hemolytic anemia (T. Necheles et al., 1968; Steinberg et al., 1970; T. F. Necheles et al., 1964). Essentially similar effects were obtained on incubating the red cells in the presence of divicine or isouramil (Mager et al., 1965). Addition of glucose obviated the deleterious effects of the drugs in normal but not in G6PD-deficient erythrocytes. The primary site of the antimetabolic action of APH was traced to hexokinase (Kosower et al., 1964; Mager et al., 1964). Furthermore, the APHinduced inhibition of hexokinase was shown to be mediated by GSSG formation, thus representing a particular case of so-called disulfide poisoning previously described by Eldjarn and Bremer (1962). While the potential significance of hexokinase inhibition in shortening the life span of the red cells need scarecely be elaborated, it remains to be seen whether this metabolic derangement plays an essential part in the actual mechanism of red cell destruction occurring in drug-induced hemolysis and favism (see Brewer et A number of enzymatic alterations in G6PD-deficient erythrocytes were re-

ported to occur independently of their exposure to noxious drugs (Schrier ef al, 1958, 1959; Larizza ef al., 1958). Observations on the decrease in the activities of NADPH-diaphorase (Jaffé, 1963), phosphomonoesterase (Oski et al., 1963; Bottini and Modiano, 1965), and pyrophosphatase (Scheuch et al., 1961; Brunetti et al., 1962a,b) appear to be of particular interest. The reduced activity of these enzymes, which are known to be SH dependent, may be due to the inclement environment created by the diminished GSH level. Furthermore, it has been shown that GSSG inhibits the activity of a variety of enzymes, such as glucose-6-phosphate dehydrogenase, inorganic pyrophosphatase, triosephosphate dehydrogenase (Scheuch and Rapoport, 1962), hexokinase (Eldjarn and Bremer, 1962; Mager et al., 1964), and ATPase (Kutscher, 1961). 289 Allison, À. C., and Clyde, D. F. (1961). Br. Med. J. 1, 1346. Allison, A. C., Charles, L. J., and McGregor, I. A. (1961). Nature (London) 190, 1198. Amin-Zaki, L., El-Din, S., and Kubba, K. (1972). Bull. W. H. O. 47, 1. Anderson, L. A. P., Howard, A., and Simonsen, J. L. (1925). Indian J. Med. Res. 12, 613. Andrews, R. S., and Pridham, J. B. (1965). Nature (London) 205, 1213. Angeloy, A, and Andrev, I. (1959). Vop. Pediat. Akus. Ginekol. No. 2, p. 7 (cited after Belsey, 1973). Arscott, G. H., and Harper, J. A. (1963). J. Nutr. 80, 251. Auquier, L., Paolaggi, J. B., and Dastugue, B. (1968). Sem. Hop. 44, 2037. Aurichio, L. (1935). Rass. Clin.-Sci. 13, 20 Belsey, M. A. (1973). Bull. W. H. O. 48, 1-13. Bendich, A., and Clements, G. C. (1953). Biochim. Biophys. Acta 12, 462. Beutler, E. (1957). J. Lab. Clin. Med. 49, 84. Beutler, E. (1970). Blood 36, 523. Beutler, E. (1971). Semin. Hematol. 8, 311. Beutler, E. (1972), In “The Metabolic Basis of Inherited Disease’ (J. B. Stanbury, J. B. Wyngaar- IV. CONCLUDING REMARKS The data reviewed in this chapter clearly indicate that, despite the considerable progress achieved in the research on favism, there are still serious gaps in our understanding of the pathogenesis of this disease. Particularly perplexing is the inadequacy of our current knowledge to account for the sporadic and rather capricious incidence of favism, as well as the absence of a clear-cut correlation between the degree of exposure of the susceptible individuals to the noxious principle of the fava plant and the occurrence of the hemolytic syndrome. It appears reasonable to surmise that the epidemiology of favism is governed not only by genetic factors, but also by a number of environmental determinants, such as variations in the amount of the toxic principle present in different varieties of the Vicia faba plant, as well as differences in the food and cooking habits of the G6PD-deficient subjects in the various populations. The major and most urgent issue, however, is the definitive establishment of the chemical identity of the causative agent of favism. It is obvious that the achievement of this goal would greatly contribute to the elucidation of other aspects of this disease and might also provide a rational basis for a prophylactic and therapeutic approach. ACKNOWLEDGMENTS Part of the original work of the authors referred to in the text and the preparation of this chapter were supported by Grant P15/181/16 from the World Health Organization and by an award from the Chief Scientists’ Office of the Ministry of Health, Israel. REFERENCES Aebi, H. E. (1967). Annu. Rev. Biochem. 36, 271. Allen, D. W., and Jand!, J. H. (1961). J. Clin. Invest. 40, 454. den, and D. B. Fredrickson, eds.), 3rd ed., p. 1358. McGraw-Hill, New York. Beutler, E., and Baluda, M. C. (1963). Blood 22, 323. Beutler, E., and Mitchell, M. (1968). Blood 32, 816. Beutler, E., Dern, R. J., Flanagan, C. L., and Alvin, A. S. (1955). J. Lab. Clin. Med. 45, 286. Beutler, E., Robson, M., and Buttenwieser, E. (1957). J. Clin. Invest. 36, 617. Beutler, E., Yeh, M., and Fairbanks, V. F. (1962) Proc. Natl. Acad. Sci. U.S.A. 48, 9. Beutler, E., Duron, O., and Kelly, B. M. (1963) J. Lab. Clin. Med. 61, 882. Bien, S., Salemnik, G., and Zamir, L. (1968). J. Chem. Soc. C p. 496. Bogair, N. (1951). Harofeh Haivri 24, 72 (in Hebrew). Boivin, P., and Galand, C. (1965). Nouv. Rev. Fr. Hematol. 5, 707. Boivin, P., Rogé, G., and Guéroult, N. (1969). Enzymol. Biol. Clin. 10, 68. Boivin, P., Hakim, J., and Blery, M. (1970). Presse Med. 78, 171. ean H., Davenport, H. W., Jeffreys, C. E. P., and Warner, R. C. (1937). J. Biol. Chem. 117, 7. Bottini, E. (1973). J. Med. Genet. 10, 154. Bottini, E., and Modiano, G. (1965). Acta Paediatr. Lat. 18, 242. Bottini, E., Lucarelli, P., Spennati, G. F., Businov, L., and Palmarino, R. (1970). Clin. Chim. Acta 30, 831. Bottini, E., Lucarelli, P., Agostino, R., Palmarino, R., Businco, L., and Antognosi, G. (1971). Science 171, 409, Bowman, J. E., and Walker, D. G. (1961). Nature (London) 189, 555. Boyer, S. H., and Graham, J. B. (1965). Am. J. Hum. Genet. 17, 320. Boyer, S. H., Porter, I. H., and Weilbacher, R. G. (1962). Proc. Natl. Acad. Sci. U.S.A. 48, 1868. Brewer, G. J., Tarlov, A. R., and Kellermeyer, R. W. (1961). J. Lab. Clin. Med. 58, 217. Brewer, G. J., Tarlov, A. R., and Alving, A. S. (1962). J. Am. Med. Assoc. 180, 386. ripa G. J., Powell, R. D., Swanson, S. H., and Alving, A. S. (1964). J. Lab. Clin. Med. 64, 01. Brunetti, P., Rosetti, R., and Broccia, G. (1960). Rass. Fisiopatol. Clin. Ter. 32, 338. Brunetti, P., Grignani, F., and Ernisli, G. (1962a). Acta Haematol. 27, 146. Brunetti, P., Grignani, F., and Ernisli, G. (1962b). Acta Haematol. 27, 246. Bunn, H. F., and Jandl, J. H. (1966) Proc. Natl. Acad. Sci. U.S.A. 56, 974. Carcassi, U. (1958). ‘‘Eritroenzimopatie e anemie emolitiche.’’ Omnia Med., Pisa. Carson, P. E. (1960). Fed. Proc., Fed. Am. Soc. Exp. Biol. 19, 995. Carson, P. E., Flanagan, C. L., Ickes, C. E., and Alving, A. S. (1956). Science 124, 484.

Chan, T. K., Todd, D., and Wong, C. C. (1964). Br. Med. J. 2, 102. Chan, T. K., Todd, D., and Wong, C. C. (1965). J. Lab. Clin. Med. 66, 937. Charlton, R. W., and Bothwell, T. H. (1959). S. Afr. J. Med. Sci. 24, 88. J. Chem. Soc. p. Chesterfield, J. H., Hurst, D. T., McOmie, J. F. W., and Tute, M. S. (1964). 1001. Chevion, M., Navok, T., Glaser, G., and Mager, J. (1979). To be published. Childs, B., Zinkham. W., Browne, E. A., Kimbro, E. L., and Torbert,.J. V. (1958). Hopkins Hosp. 102, 21. Bull. Johns Jacob, H. S. (1970), Semin. Hematol. 7, 341. Jacob, H. S., and Jandl, J. H. (1962a). J. Clin. Invest. 41, 779. Jacob, H. S., and Jandl, J. H. (1962b). J. Clin. Invest. 41, 1514. Jacob, H. S., and Jandl, J. H. (1966). N. Engl. J. Med. 274, 1162. Jacob, H. S., and Winterhalter, K. (1970). Proc. Natl. Acad. Sci. U.S.A. 65, 697 Jaffé, E. R. (1963). Blood 21, 561. | | Johannsen, L. P., Witt, I., and Künzer, W. (1968). Dsch. Med. Wochenschr. 93, 2463. Johnson, T. B. (1914). J. Am. Chem. Soc. 36, 337. Kantor, S. Z., and Arbesman, L. (1959). J. Allergy 30, 114. Chung, S. F. (1965). Paediatr. Indonesia 5, Suppl., 880. Cohen, G. (1966). Biochem. Pharmacol. 15, 1775. Cohen, G., and Hochstein, P. (1961). Science 134, 1756. Kantor, S. Z., Pinkhas, J., and Djaletti, M. (1962). J. Allergy 33, 390. Kattamis, C. A. (1969). Ann. Soc. Belg. Med. Trop. 49, 289. Cohen, G., and Hochstein, P. (1963). Biochemistry 2, 1420. Cohen, G., and Hochstein, P. (1964). Biochemistry 3, 895. Clin. Ter. 33, 42. Contu, J., Pitzus, F., Lenzerini, L., and Marcolongo, R. (1961). Rass. Fisiopatol. Creger, W. P., and Gifford, H. (1952). Blood 7, 721. Crosby, W. H. (1956). Blood 11, 91. London. Dacie, J. V. (1954). “The Hemolytic Anemias,’’ p. 354. Churchill, Davies, P. (1962). Q. J. Med. 31, 157. Davoll, J., and Laney, D. H. (1956). J. Chem. Soc. p. 2124. Dawson, J. P., Thayer, W. W., and Desforges, J. F. (1958). Blood 13, 1113. S. (1954). J. Lab. Dern, R. J., Weinstein, I. M., Leroy, G. V., Talmage, D. W., and Alving, A. Clin. Med. 43, 303. 301. Dickson, L. G., Johnson, C. B., and Johnson, D. R. (1973). Clin. Chem. 19, 513. 40, O. H. W. Bull. (1969). H. Donoso, G., Hedayat, H., and Khayatyan, Du, S. D. (1952). Chin. Med. J. 70, 17. Efrati, P. (1952). Dapim Refu. 11, 134 (in Hebrew). Eldjarn, L., and Bremer, J. (1962). Biochem. J. 84, 286. Ellman, G. L. (1959). Arch. Biochem. Biophys. 82, 70. the Preservation of Emerson, C. P., Ham, T. H., and Castle, W. B. (1949). In “Conference on Am. Natl. Red 115. p. ed.), Potthoff, J. (C. Blood”’ the of Proteins the of and Formed Elements Cross, Washington, D.C. Fairbanks, V. F., and Beutler, E. (1962) Blood 20, 591. Fermi, C., and Martinetti, P. (1905). Ann. Ig. 15, 75. Kattamis, C. A. (1971). Arch. Dis. Child. 46, 741. Kattamis, C. A., Kyriazakon, M., and Chaidas, S. (1969). J. Med. Genet. 6, 34 Kidson, C., and Gajdusek, D. C. (1962). Aust. J. Sci. 25, 61. a Kidson, C., and Gorman, J. G. (1962). Am. J. Phys. Anthropol, 20, 357. Kirkman, H. N., and Hendrickson, E. M. (1963). Am. J. Hum. Genet. 15, 241. nr Here P. R., and Naiman, I. L. (1964). Cold Spring Harbor Symp. Quant. Kleiner, I. S. (1912). J. Biol. Chem. 11, 443. Kosower, N. S., and Kosower, E. M. (1967). Nature (London) 215, 286. Kosower, N. S., Vanderhoff, G. A., and London, I. M. (1964). Nature (London) 201, 684. Kosower, N. S., Song, K. R., and Kosower, E. M. (1969). Biochim. Biophys. Acta 192 15 Kutscher, H. (1961). Folia Haematol, (Leipzig) 78, 360. Langley, G. B., Todd, F. R., and Bishop, A. J. (1969). Can. Med. Assoc. J. 100, 973. Larizza, P., Brunetti, P., Grignani, F., and Ventura, S. (1958). Haematologica 43, 205. Larizza, P., Brunetti, P., and Grignani, F. (1960). Haematologica 45, 1, 129. Lee, J. G. (1950). J. Nutr. 40, 587. | Levene, P. A. (1914). J. Biol. Chem. 18, 305. Liener, I. E. (1966). Publ. 1354, 47-50. Lin, J. Y. (1963). J. Formosan Med. Assoc. 62, 777. Lin, ae and Ling, K. H. (1962a). J. Formosan Med. Assoc. 61, 484; Chem. Abstr. 65, 4143 Lin, a and Ling, K. H. (1962b). J. Formosan Med. Assoc. 61, 490; Chem. Abstr. 65, 4144 Flohé, L., and Brand, J. (1969). Biochim. Biphys. Acta 191, 541. Flohé, L., Niebch, G., and Reiber, H. (1971). Z. Klin. Chem. Klin. Biochem. 9, 431. Flohé, L., Günzler, W. A., and Schocke, H. H. (1973). FEBS Lett, 32, 132. Francoeur, M., and Denstedt, O. F. (1954). Can. J. Biochem. Physiol. 32, 663. 26, 785. Gaetani, G., Salvidio, E., Panaciulli, I., Ajmar, F., and Paravidino, G. (1970). Experientia Gasbarrini, A. (1915). Policlinico, Sez. Prat. 22, 1505. Gehrmann, von G., Sturm, A., and Amelung, D. (1963). Dtsch. Med. 291 Wochenschr. 88, 1865. Gelpi, A. P. (1965). Blood 25, 486. 185, 257. Gilles, H. M., Watson-Williams, J., and Taylor, B. G. (1960). Nature (London) Glaser, G., Weissberg, J., Chevion, M., and Mager, J. (1979). To be published. Greenberg, M. S., and Wong, H. (1961). J. Lab. Clin. Med. 57, 733. Guggenheim, M. (1913). Z. Phys. Chem. 88, 276. Hassan, M. M. (1971). J. Trop. Med. Hyg. 74, 187. Trop. Geogr. Hedayat, Sh., Rahbar, S., Mahbooli, E., Ghaffarpour, M., and Sobhi, N. (1971). Med. 23, 149-157. Higazi, M. L., and Read, W. W. C. (1974). J. Agric. Food. Chem, 22, 570. 21, 1327. Ikeda, K., Sumi, T., Yokoi, K., and Mizuno, Y. (1973). Chem. Pharm. Bull. Lin, J. Y., and Ling, K. H. (1962c). J. Formosan Med. Assoc. 61, 579. Loehr, G. W., and Waller, H. D. (1961). Dsch. Med. Wochenschr. 86, 27. Loehr, G. W., and Waller, H. D. (1962). Med, Klin. (Munich) 36, 1521. Luisada, A. (1941). Medicine (Baltimore) 20, 229. Luzatto, L., Usanga, E. A., and Reddy, S. (1969). Science 164, 839. Lyon, M. F. (1961). Nature (London) 190, 372. McOmie, J. F. W., and Chesterfield, J. H. (1956). Chem. Ind. (London) 2, 1453. u J., Razin, A., Hershko, A., and Izak, G. (1964). Biochem. Biophys. Res. Commun. 17, 703. ee ee A., Izak, G., Bien, S., and Noam, M. (1965). Biochem. Biophys. Res. Manai, A. (1929). “Il favismo.’’ Edit. Stamp. Libreria Ital. e Stran., Sassari. Marcolongo, F., Carcassi, U., and Cusudda, U. (1950). Sci. Med. Ital. 1, 625. Marks, P. A., and Banks, J. (1965). Ann. N. Y. Acad. Sci. 123, 198. Marks, P. A., and Gross, R. T. (1959). J. Clin. Invest. 38, 2253. Marks, P. A., Gross, R. T., and Hurwitz, R. E. (1959). Nature (London) 183, 1266. Mela, C., and Perona, G. P. (1959). Boll. Soc. Ital. Biol. Sper. 35, 146.

Mills, G. C., and Randall, H. P. (1958). J. Biol. Chem. 232, 589. Minnich, V., Smith, M. B., Brauner, M. J., and Majerus, P. W. (1971). J. Clin. Invest. 50, 507. Mohler, D. N., and Williams, W. J. (1961), J. Clin. Invest. 40, 1753. Motulsky, A. G. (1960). Hum. Biol. 32, 28. Motulsky, A. G., and Campbell-Kraut, J. M. (1961). Proc. Conf. Genet. Polymorphism Geogr. Variations Dis., 1960 p. 159. Nagel, R. E., and Ranney, H. M. (1973). Semin. Hematol. 10, 269. Nathan, R. D., Pachtman, E. A., Fiorelli, G., and Frumin, A. M. (1974). Am. J. Clin. Pathol. 61, 462. Necheles, T. F., Boles, T., and Allen, D. (1968). J. Pediat. 72, 319. Necheles, T. F., Steinberg, M. H., and Cameron, D. (1970). Br. J. Haematol. 19, 605. Oort, M., Loos, J. A., and Prins, H. K. (1961). Vox Sang. 6, 370. Oski, F. A., Shahidi, N. T., and Diamond, L. K. (1963). Science 139, 409. Panizon, F. (1967). Minerva Pediatr. 19, 1391. Panizon, F., and Pujatti, G. (1957). Stud. Sassaresi 35, 105. Panizon, F., and Vullo, C. (1961). Acta Haematol. 26; 337. Panizon, F., and Vullo, C. (1962). Haematologica 47, 205. Panizon, F., and Zacchello, F. (1965). Acta Haematol. 33, 129. Panizon, F., and Zacchello, F. (1966). Arti Soc. Med.-Chir. Padova Fac. Med. Chir. Univ. Padova 41,5. Pannaciulli, I., Tizianello, A., Ajmar, F., and Salvidio, E. (1965). Blood 25, 92. Patterson, J. W., Lazarow, A., and Levey, S. (1949). J. Biol. Chem. 177, 197. Perera, C. B., and Frumin, A. M. (1965). Bibl. Haematol. 23, 589, Piomelli, S., Corash, L. M., Davenport, D. D., Miraglia, G., and Amorosi, E. L. (1968). J. Clin. Invest. 47, 940. Plato, C. C., Rucknagel, D. L., and Gershowitz, H. (1964). Am. J. Hum. Genet. 16, 267. Porter, I. H., Schulze, J., and McKusick, V. A. (1962). Ann. Hum. Genet. 26, 107. Pridham, J. B., and Saltmarsh, M. J. (1963). Biochem. J. 87, 218. Prins, H. K., and Loss, J. A. (1969). In “Biochemical Methods in Red Cell Genetics’’ (Y. Y. Yunis, ed.), p. 123. Academic Press, New York. Prins, H. K., Oort, M., Loos, J. A., Zurcher, C., and Beckers, T. (1966). Blood 27, 145. Rachmilewitz, E. A., Peisach, J., Bradley, T. B., Jr., and Blumberg, W. E. (1969). Nature (London) 222, 248. Ragab, A. H., El-Alfi, O. S., and Abboud, M. A. (1966). Am. J. Hum. Genet. 18, 21. Razin, A., Hershko, A., Glaser, G., and Mager, J. (1968). Isr. J. Med. Sci. 4, 862. Rifkind, R. A. (1965). Blood 26, 433. Rifkind, R. A. (1966). Am. J. Med. 41, 711. Ritthausen, H. (1876). Ber. Dtsch. Chem. Ges. 9, 301. Ritthausen, H. (1881). J. Prakt. Chem. [N.S.] 24(2), 202. Ritthausen, H. (1896). Ber. Disch. Chem. Ges. 29, 2108. Ritthausen, H. (1899a). J. Prakt. Chem. 59, 480. Ritthausen, H. (1899b). J. Prakt. Chem. 59, 482. Ritthausen, H., and Kreusler, U. (1870). J. Prakt. Chem. [N.S.] 2(2), 333 Rokicka-Milewska, R., Maj, S., and Jablonska-Skwiecinska, E. (1968). Pediat. Pol. 43, 621. Roth, K. L., and Frumin, A. M. (1960). J. Lab. Clin. Med. 56, 695. Rotruck, J. T., Pope, A. L., Ganther, H. E., and Hoekstra, W. C. (1972). J. Nutr. 102, 689. Rotruck, J. T., Pope, A. L., Ganther, H. E., Swanson, A. B., Hafeman, D. G., and Hoekstra, W G. (1973). Science 179, 588. u Rozynkowa, D., Gebala, A., and Zagörski, Z. (1970). Pol. Med. J. 9, 70. Rozynkowa, D., Matwijow, J., and Zagórski, Z. (1971). Genet. Pol. 12, 588. Russell, B. (1965). ‘‘History of Western Philosophy,’ 2nd ed., p. 50. Allen & Unwi, London Ryan, B. P. K., and Parsons, I. C. (1961). Nature (London) 192, 477. | | Sansone, G., and Segni, G. (1956). Boll. Soc. Ital. Biol. Sper. 32, 456. Sansone, G., and Segni, G. (1957a). Lancet 2, 295. Sansone, G., and Segni, G. (1957b). Boll. Soc. Ital. Biol. Sper. 33, 1057. Sansone, G., and Segni, G. (1958). Boll. Soc. Ital. Biol. Sper. 34, 327. , Sansone, G., Piga, A. M., and Segni, G. (1958). Il favismo. Minerva Med. Sansone, G., Rasore-Quartino, A., and Veneziano, G. (1963). Pathologica 4, 371. Say, B., Ozand, P., and Berkel, I. (1965). Acta Paediatr. Scand. 54, 319. D., and Rapoport, S. M. (1962). Acta Biol. Med. Ger. 8, 31. (London) 190. DEC., Ockel, E., Wagen! . and Rapoport, S. M. (1961). Nature D., Kahrig, Scheuch, genknecht, L., Schmitz, K., and Fritz, K. W. (1968). Med. Welt 23, 1419. Schneer, J. H. (1968). Acta Haematol. 40, 44. Schrier, S. L., Kellermeyer, R. W., and Alving, A. S. (1958). Proc. Soc. Exp. Biol. Med. 99,354. Schrier, S. L., Kellermeyer, R. W., Carson, P. E., Ickes, C. E., and Alving, A. S. (1959). 1. Lab Clin. Med. 54, 232. Schulze, E. (1891). Hoppe-Seyler’s Z. Physiol. Chem. 15, 140. | Shaker, Y., Onsi, A., and Aziz, R. (1966). Am. J. Hum. Genet. 18, 609. Siniscalco, M., Bernini, L., Latte, B., and Motulsky, A. G. (1961). Nature (London) 190, 1179 Srivastava, S. K., and Beutler, E. (1970) Biochem. J. 119, 353. | | Stamatoyannopoulos, G., Fraser, G. R., Motulsky, A. G., Fessas, P., Akrivakis, A., and Papayonnopoulou, T. (1966). Am. J. Hum. Genet. 18, 253. Steinberg, M., Brewer, M., and Necheles, T. (1970). Arch. Intern. Med. 125, 302. Symvoulidis, A., Voudiclaris, S., Mountokalakis, T., and Pougounias, H. (1972). Lancet 2, 819 Szeinberg, A., and Chari-Bitron, A. (1957). Acta Haematol. 18, 229. | | Szeinberg, A., and Sheba, C. (1960). Personal communication to Motulsky (1960). Szeinberg, A., Sheba, C., Hirshorn, N., and Bodonyi, E. (1957). Blood 12, 603. Szeinberg, A., Asher, Y., and Sheba, C. (1958a). Blood 13, 348. Szeinberg, A., Sheba, C., and Adam, A. (1958b). Nature (London) 181, 1256. Szeinberg, A., Sheba, C., and Adam, A. (1958c). Blood 13, 1043. Taleb, N., Loiselet, J., and Guorra, F. (1964). C.R. Hebd. Seances. Acad. Sci. 258, 5749. Tarlov, A. R., Brewer, G. J., Carson, P. E., and Alving, A. S. (1962). Arch. Intern. Med. 109, 209 U.S. Department of Agriculture (1963). U.S., Dep. Agric., Agric. Handb. 8. | | Vella, F. (1959). Med. J. Austr. 2, 196. Vella, F., and Ibrahim, S. A. (1962). Sudan M. J. 1, 136 (cited after Hassan, 1971). Vince-Ribarié, V. (1962). Lijech. Vjesn. 84, 151. | Waldron, H. A. (1973). Br. Med. J. 2, 667. Walker, D. G., and Bowman, J. E. (1959). Nature (London) 184, 1325. Walker, D. G., and Bowman, J. E. (1960). Proc. Soc. Exp. Biol. Med. 103, 476. Waller, H. D., and Benöhr, H. C. (1973). In ‘Erythrocytes, Thrombocytes, Leukocytes’’ (E. Gerlach, ed.), p. 200. Thieme, Stuttgart. Waller, H. D., and Gerok, W. (1964). Klin. Wochenschr. 42, 948. Waller, H. D., Loehr, G. W., Zysno, E., Gerok, W., Voss, D., and Strauss, G. (1965). Klin. Wochenschr. 43, 413. ee Mills, G. C. (1957). J. Biol. Chem. 229, 189. Mills, G. C. (1959). J. Biol. Chem. 234, 502. Mills, G. C. (1960). Arch. Biochem. Biophys. 86, 1.

25. Waller, H. D., Benoehr, H. C., and Waumans, P. (1969). Klin. Wochenschr. 47, Winterstein, A., and Somló, F. (1933). In ‘“Handbuch der Pflanzenanalyse”’ (G. Klein, ed.), Springer-Verlag, Berlin and New York. Wong, H. B. (1972). J. Singapore Paediatr. Soc. 14, 17. Yoshida, A. (1966). J. Biol. Chem. 241, 4966. p. 362. Yoshida, A. (1967a). Proc. Natl. Acad. Sci. U.S.A. 57, 838. Yoshida, A. (1967b). Biochem. Genet. 1, 81. Yoshida, A. (1970). J. Mol. Biol. 52, 483. Yoshida, A. (1973). Science 179, 532. Yoshida, A., and Hoagland, V. D., Jr. (1970). Biochem. Biophys. Res. Commun. CHAPTER 10 40, 1167. Yoshida, A., and Lin, M. (1973). Blood 41, 877. Yoshida, A., Stamatoyannopoulos, G., and Motulsky, A. G. (1967). Science 155, oe Yoshida, A., Stamatoyannopoulos, G., and Motulsky, A. G. (1968). Ann. N.Y. Acad. Sci. 155, 868. Yoshida, A., Beutler, E., and Motulsky, A. G. (1971). Bull. W. H.O. 45, 243. Zannos-Mariolea, L., and Kattamis, C. (1961). Blood 18, 34. { llervens } FRANK PERLMAN Zinkham, W. H., Lenhard, R. E., Jr., and Childs, B. (1958). Bull. Johns Hopkins Hosp. 102, 169. I. Introduction een eneen eee eeen II. Clinical Disorders .......,...,........,....,............,.,.............. A. General Consideration of Symptoms .......,...,......,,,.,,............ B. Factors Influencing Symptoms ........,,..,,.,,....,..,...,............ II. Immunological Aspects .,..................,.,,.,.,.............,....4.. 295 296 296 296 A. Genetic Control of the Immune Response .....,.,..,...,.......,..,.,.... B. Antibodies , ua... 8 ae... oS 298 299 300 IV. Antigens (Food Allergens) Li... A. General Considerations ..................,.,.,...,..,.,............... 303 303 | | B. Biological and Botanical Relationships ........,,,..,.,..........,,....4. C. Allergenic Specificity within Individual Plants .......,,,.........,........ D. Nature of Food Allergens ............................,,,.............. 303 303 304 | | | V. Detection of Food Allergens ........,.........,..,..,.,................... A. Subjective Methods... B. Objective Methods … een eeen eee VI. Specific Food Allergens …. .... …. 2.2... 02 e a 311 311 313 316 | B, Vegetables ..... a veren an oe Se TS C. Fruits eneen 318 E. Stability of Food Allergens ee A. Cereal Grains Li... iii D. Nuts, Seeds, and Beans .......,..,,.........,, eee... E. Miscellaneous Food Allergens ............,,,...........,.,,,,......... VIL, Summary iii References . ne. 4 venen RR O OA 306 316 | | | | 319 319 | | 321 324 | Î 325 | | I. INTRODUCTION Allergens hold a somewhat anomalous position in a discussion of naturally occurring food toxins. The true toxins are undesired constituents of some foods and exhibit their effects on anyone who consumes them. The severity of such toxic effects is roughly proportional to the quantity consumed. On the other hand, allergens are usually normal food constituents, and the abnormality rests in | | | | | | 295 TOXIC CONSTITUENTS OF PLANT FOODSTUFFS, SECOND EDITION Copyright © 1980 by Academic Press, Inc. All rights of reproduction in any form reserved. ISBN 0-12-449960-0