Cyanogenic Glycosides.docx

CYANOGENIC GLYCOSIDES First draft prepared by Dr G. Speijers, National Institute of Public Health and Environmental Protection Laboratory for Toxicology Bilthoven, The Netherlands 1. EXPLANATION Cyanogenic glycosides are phytotoxins which occur in at least 2000 plant species, of which a number of species are used as food in some areas of the world. Cassava and sorghum are especially important staple foods containing cyanogenic glycosides (Conn, 1979a,b; Nartey, 1980; Oke,1979, 1980; Vennesland et al., 1982; Rosling, 1987). There are approximately 25 cyanogenic glycosides known. The major cyanogenic glycosides found in the edible parts of plants used for human or animal consumption are summarized in Table 1. The potential toxicity of a cyanogenic plant depends primarily on the potential that its consumption will produce a concentration of HCN that is toxic to exposed animals or humans (see Table 2). Several factors are important in this toxicity: The first aspect is the processing of plant products containing cyanogenic glycosides. When the edible parts of the plants are macerated, the catabolic intracellular enzyme ß-glucosidase can be released, coming into contact with the glycosides. This enzyme hydrolyzes the cyanogenic glycosides to produce hydrogen cyanide and glucose and ketones or benzaldehyde. The hydrogen cyanide is the major toxic compound causing the toxic effects. Plant products (notably cassava), if not adequately detoxified during the processing or preparation of the food, are toxic because of the release of this preformed hydrogen cyanide. The second aspect is the direct consumption of the cyanogenic plant. Maceration of edible parts of the plants as they are eaten can release ß-glucosidase. The ß-glucosidase is then active until the low pH in the stomach deactivates the enzyme. Additionally, it is possible that part of the enzyme fraction can become reactivated in the alkaline environment of the gut. At least part of the potential hydrogen cyanide is released, and may be responsible for all or part of the toxic effect of cyanogenic glycosides in the cases of some foods. Table 1: The occurrence of cyanogenic glycosides in major edible plants (Conn, 1979a,b) Cyanogenic glycosides

Plant species Common name

Latin name

Amygdalin

almonds

Prunus amygdalus.

Dhurrin

sorghum

Sorghum album, Sorghum bicolor.

Linamarin

cassava

Manihot esculenta, M. carthaginensis Phaseolus lunatus.

lima beans Lotaustralin

cassava lima beans

Manihot carthaginensis Phaseolus lunatus.

Prunasin

stone fruits

Prunus species e.g., P. avium, P. padus, P. persica, P.

bamboo shoots

Bambusa vulgaris.

macrophylla. Taxiphyllin

The third aspect is that the cyanogenic glycosides taken up intact with the food are (partly) hydrolyzed by the ß-glucosidase activity of the bacteria of the gut flora of animals or humans (Conn, 1979a,b; Oke, 1979, 1980; Nartey, 1980; Rosling, 1987; Gonzales & Sabatini, 1989). Cyanide, released from a cyanogenic glycoside in food by ß-glucosidase either of plant or from gut microflora origin and taken up, follows the known cyanide metabolic pathway and toxicokinetics both for animals and man. Cyanide is detoxified by the enzyme rhodanese, forming thiocyanate, which is excreted by urine (Conn, 1979a,b; Oke, 1979, 1980). Due to several factors influencing hydrolysis of cyanogenic glycosides and the confounding influence of nutritional status (such as riboflavin, vit. B12, sodium, methionine intake) human case studies and epidemiological studies of the chronic toxicological effects have shown very variable results and were not conclusive. In addition, the data in these studies are rarely of a quantitative character (Conn, 1979a,b; Oke, 1979, 1980; Nartey, 1980; Rosling, 1987). In several studies both in animals and man the toxicity of cyanogenic glycosides is often expressed as mg releasable cyanide. Table 2: Concentration of cyanide in some tropical foodstuffs (Summarized in Nartey, 1980). Plant/tissue Cassava(bitter)/dried root cortex Cassava(bitter)/leaves Cassava(bitter)/whole tubers Cassava(sweet)/leaves Cassava(sweet)/whole tubers Sorghum/whole immature plant

mg HCN/kg 2450 310 395 468 462 2500

Bamboo/immature Lima beans from Lima beans from Lima beans from

shoot tip Java (coloured) Puerta Rico (black) Burma (white)

8000 3120 3000 2100

Because this monograph first discusses the toxicity of cyanide as a basis for understanding that of cyanogenic glycosides, a modified form of the general monograph format has been used, presenting first biological data for cyanide, then that for cyanogenic glycosides. CYANIDE 2. 2.1

BIOLOGICAL DATA Biochemical aspects

2.1.1.1

Absorption, distribution, and excretion

Hydrogen cyanide after oral administration is readily absorbed (it is also readily absorbed after inhalation exposure and through skin and eyes). After absorption, cyanide is rapidly distributed in the body through the blood. The concentration of cyanide is higher in erythrocytes than in plasma. It is known to combine with iron in both methaemoglobin and haemoglobin present in erythrocytes. The cyanide level in different human tissues in a fatal case of HCN poisoning has been reported: gastric content; 0.03, blood; 0.50, liver; 0.03, kidney; 0.11, brain; 0.07, and urine; 0.20 (mg/100 g) (EPA, 1990). The pharmacokinetics of 14CN- and S14CN- in rats exposed to these agents in diet for 3 weeks was investigated. All tissues contained radioactivity 9 h after intraperitoneal injection of 14CN-; highest radioactivity was found in the stomach (18%). Eighty per cent of this activity was in the form of thiocyanate. At this point 25% of the dose had already been eliminated in the urine and 4% in the expired air. When S14CN- was given per os to rats with elevated plasma thiocyanate levels due to chronic oral exposure to cyanide, most of the activity was eliminated in the urine and only small amounts were found in the faeces. This indicated the existence of a gastrointestinal circulation of thiocyanate (Okoh & Pitt, 1981). The excretion of an acute oral dose of 14C-labelled cyanide in urine, faeces and expired air was studied in rats (12 animals/group) pretreated orally for 6 weeks with unlabelled KCN or a control diet. Urinary excretion was the main route of elimination of 14C-labelled cyanide in these rats, accounting for 83% of the total excreted radioactivity at 12 h and 89% of the total excreted radioactivity at 24 h. The major metabolite of cyanide excreted in urine was thiocyanate, and this metabolite accounted for 71% and 79% of the total urinary activity at 12 h and 24 h, respectively. Only 4% of the mean total activity excreted was found in expired air after 12 h, and this value did not change after 24 h. Of the total activity in expired air in 24 h, 90% was present as carbon dioxide

and 9% as cyanide. When these results were compared with those observed for control rats, it was clear that the mode of elimination of cyanide carbon was altered in neither urine nor breath by the chronic intake of cyanide (Okoh, 1983). Golden hamsters exposed to cyanide by subcutaneous infusion appeared to excrete only a relatively low percentage (10-15%) of the dose as thiocyanate in the urine (Doherty et al., 1982). The major defence of the body to counter the toxic effects of cyanide is its conversion to thiocyanate mediated by the enzyme rhodanese. The conversion of cyanide to the less toxic thiocyanate by rhodanese was discovered by Lang (1933). Thiosulfate and 3-mercapto-pyruvate can act as sulfur donors, but neither free cystine nor cysteine can. The enzyme contains an active disulfide group which reacts with the thiosulfate and cyanide. The trivial name rhodanese is more widely used than that assigned by the Enzyme Commission [thiosulfate-cyanide sulfur transferase, EC. 1.8.1.1]; it has been inappropriately called rhodanase in several reports. The rhodanese-catalyzed irreversible conversion of cyanide to thiocyanate, in the presence of thiosulfate, provides a means for the treatment of cyanide poisoning. Since the enzyme, which is usually localized in the mitochondria in different tissues, is relatively abundant, but in sites which are not readily accessible to thiosulfate, the limiting factor for the conversion of cyanide is thus thiosulfate (EPA, 1990). The overall rate of in vivo detoxification of cyanide may be influenced by several minor reactions. Cystine may directly react with cyanide to form 2-imino-thiazolidine-4-carboxylic acid which is excreted in saliva and urine. Traces of hydrogen cyanide may be found in expired air, saliva, sweat and urine. A minor amount may be converted into formic acid which may be excreted in urine or participate in the metabolism of one carbon compound. One minor detoxification route is the combination of cyanide with hydroxycobalamine (vitamin B12) to form cyanocobalamine which is excreted in urine and bile. It may be reabsorbed by the intrinsic factor mechanism at the level of the ileum allowing effective recirculation of vitamin B12. Methaemoglobin effectively competes with cytochrome oxidase for cyanide and its formation from haemoglobin, effected by sodium nitrite or amylnitrite, is exploited in the treatment of cyanide (EPA, 1990). It has been reported that other species have lower rhodanese activity than the rat and hence the rat may be able to convert cyanide to thiocyanate more easily than other species (Himwich & Saunders, 1948). 2.1.2

Biotransformation No information available.

2.1.3

Effects on enzymes and other biochemical parameters

Cyanide causes a decrease in the utilization of oxygen in the tissues, producing a state of histotoxic anoxia. This occurs through

inactivation of tissue cytochrome oxidase by cyanide, which combines with Fe3+/Fe2+ contained in the enzyme. The enzyme-cyanide complex dissociation constant has been found to be 1 * 10-6 and 1 * 10-4 (moles/l) for the oxidized and reduced form of the enzyme, respectively. Thus, the affinity of cyanide for the oxidized form of the enzyme is two orders of magnitude higher than for the reduced form. However, the rate of reaction of cyanide with the reduced enzyme is twice the rate of reaction with the oxidized form. Cyanide can inhibit several other metalloenzymes most of which contain iron, copper or molybdenum (e.g., alkaline phosphatase, carbonic anhydrase), as well as enzymes containing Schiff base inter-mediates (e.g., 2-keto-4-hydroxyglutarate aldolase). The effect of sublethal doses of cyanide on the metabolism of glucose in mice has been studied using radiorespirometric techniques (Solomonson, 1981). Cyanide causes an increase in blood glucose and lactic acid levels and a decrease in the ATP/ADP ratio indicating a shift from aerobic to anaerobic metabolism. Cyanide apparently activates glycogenolysis and shunts glucose to the pentose phosphate pathway decreasing the rate of glycolysis and inhibiting the tricarboxylic acid cycle (EPA, 1990). 2.2

Toxicological studies

2.2.1

Acute toxicity studies

Lethal doses of HCN in mg/kg bw were reported for mouse, 3.7; dog, 4.0; cat, 2.0 and for cattle and sheep 2.0 (Summarized by Conn, 1979a). Table 3. Acute toxicity of cyanide Acute toxicity Species

Route

LD50 (mg/kg bw)

Mouse Rat Guinea-pig Rabbit Cat Dog Monkey

i.v. i.v. i.v. i.v. i.v. i.v. i.v.

0.99 0.81 1.43 0.66 0.81 1.34 1.30

Mouse

s.c. i.v. oral i.v. oral

6.0 (KCN) 2.5 (KCN) 10-15 (KCN) 2.5 (KCN) 5.3 (KCN)

Rat Dog

(HCN) (HCN) (HCN) (HCN) (HCN) (HCN) (HCN)

References

EPA EPA EPA EPA EPA EPA EPA

1990* 1990* 1990* 1990* 1990* 1990* 1990*

WHO, WHO, WHO, WHO, WHO,

1965* 1965* 1965* 1965* 1965*

Rabbit Guinea-pig Dog

s.c. s.c. i.v.

2.2 (NaCN) 5.8 (NaCN) 2.8 (NaCN)

WHO, 1965* WHO, 1965* WHO, 1965*

Mouse

oral i.v. oral i.p.

598 484 765 540

WHO, WHO, WHO, WHO,

Rat

(NaSCN) (NaSCN) (NaSCN) (NaSCN)

1965* 1965* 1965* 1965*

* as summarized in 2.2.2 2.2.2.1

Short-term toxicity studies Rats

In a 13-week toxicity study male Sprague-Dawley rats (approximately 30 rats/group) were administered KCN in the drinking water. The dose levels were 40, 80 and 160/140 mg KCN/kg bw/24 h. Three control groups were used, respectively a normal drinking water ad libitum, a "paired drinking" group (parallel to the high dose level KCN) and a group receiving drinking water with 10% ethyl alcohol. In addition, one group received drinking water with KCN (80 mg/kg bw) and 10% alcohol. Behaviour, external appearance, body weight, food consumption (daily) and drinking water consumption (twice weekly) were recorded frequently. Extensive haematological, clinical chemical (in serum) and urine analyses were carried out in 5 animals per group in week 6 and week 13. Autopsy and macroscopy were performed after 13 weeks (approximately 20 animals/group) and 11 organs were weighed. Histopathological examination was performed in brain, kidneys, heart, liver and testes of these animals. In addition thyroids of the control, the "pair drinking" control and the high-dose group (160/140 mg/kg bw/day) were examined. There was a clear indication that reduced food consumption and body weight in the KCN groups were caused by a decrease in water consumption due to decreased palatibility. Urinalyses revealed a higher level of protein in the animals receiving KCN. The amounts of protein determined showed a clear correlation to the increasing doses of KCN, as did the drinking water. Several changes in absolute organ weights were seen in the 160/140 mg KCN/kg bw/day group. Relative weights of organs were very slightly increased in the 40, slightly increased in the 80 mg and clearly increased in the 160/140 mg KCN/kg bw groups. The thymus weight was, however, reduced in the high-dose group. Histopathological examination revealed no indication of damage to the brain, heart, liver, testes, thyroids nor kidneys due to treatment with KCN (Leuschner et al., 1989b). 2.2.3

Long-term carcinogenicity studies

No data on the carcinogenicity of hydrogen cyanide have been published. However, anticarcinogenic effects of cyanide have been reported. Longevity of mice with transplanted Ehrlich ascites

tumours and Sarcoma 180 was increased 20 to 70% on i.p. injection of sodium cyanide in the dose range 0.75 to 2.0 mg/kg bw (EPA, 1990). 2.2.4 2.2.4.1

Reproduction studies Rats

A short-term reproductive study (49 day study in adults and 28 day study in pups) was performed to evaluate the cumulative effects of adding 500 mg KCN/kg to cassava root flour-based diet in pregnant rats. This meal was prepared from a low-HCN cassava variety (21 mg HCN/kg feed). High dietary level of KCN did not gestation and lactation performance of effect of high cyanide-containing diet observed on lactation performance. The

have any marked effect in female rats. No carry-over fed during gestation was high cyanide-containing diet,

however, significantly reduced feed consumption and daily growth rate of the offspring when fed during post-weaning period. Protein efficiency ratio was not only reduced by the cyanide diet during post-weaning growth phase but there was an additional carry-over effect from gestation. Serum thiocyanate was significantly increased in lactating rats and their offspring during lactation and in the postweaning growth phase of the pups. No apparent carry-over effect was noticed on this parameter. Rhodanese activity in liver and kidneys was unaffected by feeding the high cyanide diet during gestation, lactation, nor during postweaning growth (Tewe & Maner, 1981b). 2.2.5 2.2.5.1

Special studies on embryotoxicity and teratogenicity Hamsters

Pregnant golden hamsters were exposed to sodium cyanide on days 6-9 of gestation by infusion via subcutaneously implanted osmotic minipumps. Cyanide (0.126-0.1295 mmol/kg/h) induced high incidences of resorptions and malformations in the offspring. The most common abnormalities observed were neural tube defects (Doherty et al., 1982). 2.2.6 2.2.6.1

Special studies on the thyroid gland Rats

A group of 10 male rats was fed a 10% casein diet containing added methionine, vitamin B12, iodine and potassium cyanide (1500 mg/kg feed) for nearly one year. Compared to a control group not receiving cyanide, depression of body-weight was observed throughout the study period, but there were no deaths nor clinical signs of toxicity. Depression of both plasma thyroxine and thyroxine secretion rate suggestive of depressed thyroid function were evident at 4 months but less so after 1 year. At autopsy the animals were found to have enlarged thyroids and this may have been the mechanism of adaptation. Some differences in the histopathology of the spinal cord, notably the white matter, were also found between controls and cyanide-treated animals (Philbrick et al., 1979).

2.2.6.2

Pigs

Performance and metabolic and pathological changes were evaluated in 48 growing pigs fed different levels of dietary protein (9 and 16%), cyanide, and iodine (0 and 0.36 mg iodine/kg feed) during 56 days. Protein deficiency reduced urinary iodine excretion and the concentrations of protein, protein-bound iodine (PBI) and thiocyanate in serum. It also reduced liver rhodanese activity and caused a decrease in urinary thiocyanate excretion which was not significant. Dietary cyanide increased urinary thiocyanate and iodine excretion and serum PBI. Pathological studies showed that cyanide treatment had no marked effect on the microanatomy of the tissues examined. Dietary iodine deficiency caused histological changes in the thyroid gland and bone which suggested a decline in metabolic activity. Iodine deficiency caused hyperplastic goitre in the experimental animals (Tewe & Maner, 1980). 2.2.7

Genotoxicity

Two negative and one marginally positive genotoxicity studies for cyanide have been reported. Potassium cyanide was not mutagenic in Salmonella typhimurium strains TA1535, TA1537, TA1538, TA98 nor TA100 with or without S-9 liver microsomes. Cyanide was also negative in recombinant-assay in Bacillus subtilis. One study reported marginally mutagenic activity of HCN to Salmonella typhimurium strain TA100 in the absence of S-9 mix. In the same study no mutagenic activity in strain TA98 with or without S-9 mix was observed (EPA, 1990). An in vitro Ames test with HCN in Salmonella strains TA1537, TA1538 and TA98 for detection of frame shift mutation and TA1535 and TA100 for base-pair substitutions was performed with and without metabolic activation of a S-9 microsome mixture. There was no indication of mutagenic properties under these conditions (Leuschner et al., 1983a). An in vivo mutagenicity study in Chinese hamsters detecting chromosomal aberrations with HCN orally administered to Chinese hamsters was carried out. Preparations of metaphase cells were studied for structural chromosome aberrations after 6, 24 and 48 h after oral administration of 0.4 mg HCN/kg bw. The incidence of aberrations or gaps was within the spontaneous range. Neither multiple aberrations nor pulverised metaphases were found. There was no indication of mutagenic properties relative to structural chromatid or chromosome damage (Leuschner et al., 1983b). A gene mutation assay in cultured Chinese hamster cells, V79 (genetic marker HGPRT) both in the presence and absence of metabolic activation system was carried out with KCN. The duration of the exposure with the test substance was 24 h in the experiments without S-9 mix and 2 h in the experiments with S-9 mix. The test compound dose levels employed were chosen following a preliminary toxicity experiment. The dose-levels for the main study were 400, 800, 1000, 2000 and 3000 µg/ml without S9 mix and 1000, 2000, 3000, 4000, 6000, 8000 and 10 000 µg/ml with S9 mix. KCN was tested up to a high

cytotoxicity in the absence and presence of metabolic activation. Under the present test conditions KCN was negative in the V79 mammalian cell mutagenicity test (Leuschner et al., 1989a). 2.2.8

Special studies on nervous system

A special study on the behavioural effects of chronic sublethal dietary cyanide (KCN; 0.4, 0.7 and 1.2 CN-/kg bw) in juvenile swine, mimicked the situation of free CN- intake in Liberia due to eating cassava-based foods. There were two clear behavioural trends: 1) increasing ambivalence and slower response time in reacting to various stimuli and 2) an energy conservation gradient influencing which specific behaviours would be modified in treated animals. Serum SCN- was positively correlated with daily CN- intakes. CN- treatment diminished T3 and T4 levels but elevated fasting blood glucose values (Collier-Jackson, 1988). Neuronal lesions in several animal species have been produced by chronic cyanide intoxication either by injection of unbuffered alkaline cyanide salts or by inhalation of hydrogen cyanide. The neuropathological changes include areas of focal necrosis especially around the centrum ovale, corpus striatum, corpus callosum, substantia nigra, anterior horn cells, and patchy demyelination in the periven-tricular region. In some species, the earliest effects may be on the oligodendroglia and hence myelin lesions may precede neuronal damage. Bass (1968) showed that in rats chronic cyanide intoxication produces myelin loss by its primary effect on glial cells followed by breakdown of myelin. Brierly et al. (1977) reported myelin damage and changes in the oligodendroglia in cyanide poisoning in rats. Clark (1936) described fatty degeneration of the liver and secretory tubules of the kidney in rats subjected to chronic cyanide intoxication. In these animal experiments relatively large doses of cyanide were given, often sufficient to cause partial asphyxia. Therefore it was doubtful whether the neuropathological effects demonstrated were due to asphyxia or chronic cyanide intoxication so they did not appear to have an obvious parallel to human exposure at lower levels. It is noteworthy, however, that changes in optic nerves and tracts occur consistently only in primates (Ferraro, 1933; Hurst, 1940; Lessell, 1971). In a carefully controlled experiment in which small weekly doses of cyanide were administered over several months to rats, neuronal degeneration and demyelination were reported (Smith et al., 1963). Williams and Osuntokun (1969) found that the demyelination of peripheral nerves induced in rodents by cyanide injection bore a striking resemblance to the lesions found in biopsy specimen of peripheral nerves of Nigerian patients who suffer from tropical neuropathy (reviewed by Osuntokun, 1981). 2.3 2.3.1

Observations in humans Acute toxicity studies in humans

The acute oral lethal dose of HCN for human beings is reported to be 0.5-3.5 mg/kg bw corresponding to 1.0-7.0 mg/kg bw of KCN. The clinical signs are well described (Montgomery, 1969; Gosselin et al., 1976) and include headache, dizziness, mental confusion,

stupor, cyanosis with twitching and convulsions, followed by terminal coma (Conn, 1979a). The acute oral lethal dose of HCN for man was reported to be 60 mg (Sinclair & Jeliffe, 1961). For man the acute oral dose of HCN is usually given as 50-90 mg and for potassium cyanide as 200 mg, corresponding to 81 and 110 mg HCN respectively (Lehman, 1959). Data on the oral lethal dose of cyanide for man in four cases of suicide, calculated from the amount of HCN absorbed in the body at the time of death, and from the amount of HCN found in the digestive tract, differed considerably ((calculated as mg HCN: 1450 (62.5 kg bw), 556.5 (74.5 kg), 296.7 (50.7 kg) and 29.8 (51 kg)) (Geitler & Baine, 1938). This corresponds to doses varying from 0.58-22 mg/kg bw (in WHO, 1965). CYANOGENIC GLYCOSIDES 2.

Biological data

2.1 2.1.1

Biochemical aspects Absorption, distribution, and excretion

Wistar rats (4 animals/sex/group) were given 50 mg amygdalin/rat respectively, intravenously and orally after an overnight fast. During the experiment and the night before they were kept in metabolic cages which allowed collection of urine without faeces and minimized coprophagy. After intravenous and oral administration fractions of the dose excreted by the rat as unchanged amygdalin were 70% and 0.8%, respectively. The fraction excreted as prunasin after intravenous administration was 6.6%, whereas it was 39% of the total dose after oral administration (Rauws et al., 1982). In a toxicokinetic study 10 male rats (100-120 g) were given 50 mg pure linamarin (dissolved in water, volume 0.5 ml) and 6 male rats were given water alone by stomach tube. Seven rats dosed with 50 mg linamarin died within 4 h. In a second trial 6 male rats were given 30 mg linamarin. Following dosing, urine and faeces were collected after 24, 48 and 72 h, and heparinized blood samples were taken from the optic vein or lateral tail vein. Blood was taken after 30, 40, 60, 80 and 100 min and at 2, 4, 8, 24 and 48 h. No intact linamarin was detected in faeces nor blood of the rats dosed with 30 mg (300 mg/kg bw). No linamarin in the faeces of the three surviving rats dosed with 50 mg (500 mg/kg bw) was detected, however blood and urine were not examined in these animals. Linamarin was excreted in the urine at a level of 5.65 mg (cumulated after 72 h) along with 0.823 mg of thiocyanate. These findings indicate that linamarin was absorbed intact in considerable proportion and was partially hydrolyzed (Barrett et al., 1977). In a toxicokinetic study beagle dogs (4 animals/sex/group) were administered 500 mg amygdalin in 10 ml solution respectively, intravenously and orally after overnight fast. Blood was sampled from jugular vein and urine was collected by a funnel. Faeces were

removed from the funnel. The major part of the dose (71%) was recovered in the urines collected during 6 h following intravenous amygdalin administration. The fraction of the dose excreted by glomerular filtration was calculated using the ratio of diatrizoate (which was administered simultaneously) clearance to amygdalin clearance, showing that 97% of the amount of amygdalin to be expected was recovered from the urine. The result of the experiments after intravenous administration were analyzed assuming a two-compartment model. The distribution T´ alpha was 0.10 and the elimination T´ ß 0.57. No prunasin was detected in urine (detection limit = 0.2% of dose). After oral administration of amygdalin a very low maximal plasma level is found after approximately 0.75 h. Only 2.3% of the amygdalin was systematically available (absolute bioavailablity). Prunasin was found in plasma and urine of dogs. In the urine collected during 6 h following amygdalin administration only about 1% of the dose was recovered unchanged and 21% of the dose was identified as prunasin (Rauws et al., 1982). In a toxicokinetic study prunasin was administered intravenously and orally as 100 mg doses to female dogs (2/group; 10 kg) after an overnight fast. Blood was sampled from jugular vein and urine was collected by a funnel. The results of the experiment after intravenous administration were analyzed assuming a two-compartment model. The prunasin results were compared to those obtained with amygdalin in a earlier experiment. The distribution T´ alpha was 0.08 h and elimination T´ ß was approximately 0.64 h. Prunasin was absorbed to a large extent after oral administration. The absolute bioavailability after oral administration was 50% of the administered dose. The volume of distribution (0.34 L/kg) and the clearance (0.55 L/kg h) are larger than those of amygdalin (respectively 0.19 L/kg and 0.39 L/kg h). The oral bioavailability of prunasin is considerably greater, whereas amygdalin is only slightly (2.3%) absorbed (Rauws et al., 1983). In a toxicokinetic study pure linamarin, at a dose level of 300 mg/kg bw, was administered in food to a group of Wistar rats maintained on vitamin B2-deficient, -sufficient, and -excess diets for 5 weeks and to another group of kwashiorkor rats. Free and total cyanide, intact linamarin and thiocyanate levels were estimated in urine and faeces obtained at 0, 24, 48 and 72 h periods and in blood samples obtained 72 h after the compound had been administered. There was no detectable cyanide nor intact linamarin in the faecal samples. Rats on vitamin B2-sufficient and B2-excess diets excreted higher total and free cyanide in urine than the respective vitamin B2-deficient groups. Most of the linamarin was degraded after 24 h. The rate of breakdown of the glycoside within the first 24 h was slowest for zero and half normal vitamin B2 status rats as evidenced by appearance of the glycoside in large quantities in the urine. The kwashiorkor rats, on the other hand, excreted less thiocyanate than

the controls. In addition, their control group excreted most of the thiocyanate in the first 24 h, whilst the kwashiorkor rats excreted most of the thiocyanate in the first 48 h. Dietary protein deficiency prolongs the time of metabolism and hence increases the toxicity of cyanogenic glycosides in the body (Umoh et al., 1986). 2.1.2

Biotransformation

Strained ruminal fluid was collected from cattle fed five diets (concentrate diet, freshly harvested alfalfa, cubed alfalfa, alfalfa hay and orchard grass) to determine in vitro rates of cyanogenesis from the glycosides amygdalin, prunasin and linamarin. Rates of dissociation for the corresponding aglycones, benzaldehyde cyanohydrin and acetone cyanohydrin were also determined. Hydrogen cyanide (HCN) in ruminal fluid was determined with a modified method of HCN analysis that independently measured the overall rate of cyanogenesis and the non-enzymatic dissociation of cyanohydrins, the intermediate products in the degradation of cyanogenic glycosides to HCN. Rate of dissociation of cyanohydrins in ruminal fluid was pH-dependent, with high rates of dissociation (as expressed by the rate constant or half-life of reaction) occurring at pH >6 and slower rates at pH 5 to 6. Cyanohydrin dissociation was most rapid when cattle were fasted for 24 to 48 h and ruminal pH was high; rate of dissociation was much slower during feeding and digestion. When the glycosides were examined, highest rates of cyanogenesis (mg HCN/L/s) were observed after a 24 h postprandial period. The rates were highest after feeding hay: 0.019 for amygdalin, 0.033 for linamarin and 0.048 for prunasin. Hence cattle are most susceptible to poisoning by cyanogenic plants when the pH of ruminal fluid is elevated, leading to rapid dissociation, and also when the activity of ß-glucosidase is adequate for rapid hydrolysis of glycosidic bonds. Rates of cyanogenesis were higher when ruminal inocula were from cattle fed fresh alfalfa or cubed alfalfa hay rather than from those fed grain or long hay. Rates of HCN production were lowest using inocula from cattle fed grain; rates for the three glycosides were negligible at the 3 and 6 h postprandial sampling times. In agreement with previous studies, prunasin was degraded in ruminal fluid much more rapidly than linamarin or amygdalin (Majak et al., 1989). In a comparative metabolism study rates of cyanide liberation resulting from hydrolysis of the cyanogenic glycosides linamarin, amygdalin and prunasin by a crude ß-glucosidase prepared from hamster caecum were studied in vitro. In addition, hamster blood cyanide and thiocyanate concentrations were determined at 0.5, 1, 2, 3, and 4 h after oral dose of 0.44 mmol linamarin or amydalin/kg bw. Plots of cyanide liberated versus time for linamarin and prunasin yielded straight lines, whereas for amygdalin the plot was curvilinear; the rate of cyanide release increased with time. At 10-3 M substrate concentrations, the averaged rates of hydrolysis of prunasin, amygdalin and linamarin were 1.39, 0.57 and 0.13 nmol/min/mg protein, respectively. Lineweaver-Burk plots yielded apparent Km and Vmax values of 3.63 * 10-5 M and 0.13 nmol/min/mg protein, respectively for amygdalin, and 7.33 *

10-3 M and 1.04 nmol/min/mg protein, respectively for linamarin. Blood cyanide concentrations following amygdalin treatment of the hamster reached their highest level (130 nmol/ml) 1 h after dosing and remained elevated until 3 h after treatment. Blood cyanide concentrations following linamarin treatment reached their highest level (116 nmol/ml) after 3 h and then declined immediately. Area under the blood cyanide concentration-time curve was 395 nmol h/ml for amygdalin and 318 nmol h/ml for linamarin. The results suggest a faster rate of enzymatic hydrolysis and cyanide absorption for amygdalin than for linamarin (Frakes & Sharma, 1986). In humans, pharmacological studies have shown that amygdalin is broken down to HCN, benzaldehyde and glucose by enzymes found in gut bacteria, but not intracellularly in humans. Animal and human tissues contain no significant concentrations of ß-glucosidase, the only known activating enzyme of hydrolysis of cyanogenic glycosides in vivo (Dorr & Paxinas, 1978). Cyanogenic glycosides are hydrolyzed by ß-glucosidase produced by intestinal bacteria to glucose, HCN and benzaldehyde or acetone. Benzaldehyde is oxidized to benzoic acid (and subsequently to hippuric acid) or salicylic acid isomers. Thiocyanate is present in body fluids; blood, urine, saliva, sweat and tears (Oke, 1979). The cyanide-yielding capacity of insufficiently processed cassava probably occurs as linamarin, or an intermediate break-down product, from which cyanide may be yielded in the gut by action of microbial enzymes. Significant amounts of linamarin are observed in the urine after consumption of insufficiently processed cassava as well as after consumption of other plants containing linamarin. These results indicate that linamarin, if not metabolized in the gut, will be absorbed and excreted in the urine without causing exposure to HCN. About 80% of ingested cyanide will be turned into thiocyanate and is excreted in the urine after a short period (Rosling, 1987). 2.1.3

Effects on enzymes and other biochemical parameters No information available.

2.2

Toxicological studies

2.2.1

Acute toxicity studies

Table 4. Acute toxicity studies of cyanogenic glycosides or cyanogenic plant tissues Acute toxicity Species Mouse

Sex ?

Route

LD50 (mg/kg bw)

i.p.

0.1 mmole

1981 amygdalin/kg

References Solomonson,

Rat

?

i.v.

20 000 linamarin

Oke, 1979

Rat

?

oral

450 linamarin

Oke, 1979

A dose of 25 mg linamarin (250 mg/kg bw) fed to rats (100-120 g bw) caused clinical signs of toxicity, including apnoea, ataxia and paresis. These symptoms were very marked in the absence of methionine supplementation, 50% of these rats died within 4 h. In the presence of adequate methionine supplementation, 10% of rats died and about 40% showed no signs of toxicity. The activity of Na+K+-dependent ATPase was reduced in much the same way as it was by the glycoside, digitalis (reviewed by Oke, 1980). In a toxicokinetic study 7 out 10 rats (100 g bw) died after administration of 50 mg linamarin by stomach tube (Barrett et al., 1977). Oral doses of 100, 120 and 140 mg linamarin/kg bw given by stomach tube to hamsters (90 g bw) produced signs of cyanide intoxication in a large percentage. The signs appeared within 1 h after dosing included dyspnoea, hyperpnoea, ataxia, tremors and hypothermia. Two animals dosed with 140 mg/kg bw and one animal dosed with 120 mg/kg bw died within 2 h of dosing. The signs of poisoning were greatly reduced or gone within 3 h after treatment in the surviving animals. No relationship between length of intoxication and dose was observed (Frakes et al., 1985). Hamsters have been reported to be more susceptible than rats to the acute toxic effects of orally administered amygdalin and prunasin (Willhite, 1982). A species difference in reaction to cyanogenic glycosides is observed due to a difference of detoxifying ability due to anatomical structure. Ruminants, e.g., cattle and sheep are supposed to be more susceptible to the acute toxic effects because of their larger flora of microorganisms and considerable quantities of the enzyme emulsine which hydrolyzes the glycoside (Oke, 1979). 2.2.2 2.2.2.1

Short-term toxicity studies Rats

Albino female rats (10 animals/group) were fed ad libitum one of the following diets, A; a normal laboratory diet (control), B; a 50% gari diet (Nigerian preparation of cassava), C; a raw cassava diet, D; a diet containing 5 g KCN/100 g and E; a diet containing 10 g KCN/100 g during a 14-day period. The 50% gari diet caused no significant biochemical nor haematological changes in the female rats, whereas for both the raw cassava diet and the KCN diets a decrease of Hb, PCV, total serum protein concentration and T4 concentration was observed. In the 50% gari diet group, and to a greater extent, in the other treatment groups, the serum thiocyanate levels were increased. The body weight gain was not significantly

decreased in the 50% gari group, whereas the other treatment groups showed instead of gain a loss of body weight (Olusi et al., 1979). 2.2.2.2

Guinea-pigs

In a 24-day toxicity experiment guinea-pigs (8 animals/group) were dosed daily with, respectively, laetrile (10 mg amygdalin) and 8 mg KCN/kg bw with and without ascorbic acid (100 mg). No significant effect on the body weight nor liver weight was observed. However, treatment with laetrile alone for 4, 16 or 24 days, respectively, resulted in a significant increase in urinary levels of thiocyanate. The increase was less in animals treated with vitamin C. In guinea-pigs treated with 8 mg KCN/kg bw, toxic effects were seen as evidenced in slight tremors in 3 of the 8 animals, which recovered within 5 min. All animals in the KCN group which were supplemented with ascorbate showed severe tremors, motor ataxia, bizarre neuromuscular manifestations and rhythmic head movements. The toxicity of KCN increased with elevation of vitamin C, whereas urinary excretion of thiocyanate decreased (Basu, 1983). 2.2.2.3

Chickens

In two feeding experiments (respectively 63 and 56 days) one day old broiler chickens (male and female) were fed a diet containing 0, 10, 20 or 30% cassava, respectively. The animals were studied for haematological and histopathological effects. The cassava diet studied in the 1st experiment consisted of a high-cyanide-containing cassava root meal (CRM) supplying 300 mg of total cyanide/kg, most of it in the form of cyanogenic glycosides. The cassava diet in the 2nd experiment also contained cassava foliage meal (CFM) supplying 156 mg total cyanide/kg. In the 1st experiment 26 chickens per group were used and in the 2nd experiment 160 chickens were used for the cassava groups and 80 for the control group. No changes in the haematological parameters due to cassava were seen. Addition of up to 30% CRM failed to adversely affect broiler survival, performance nor feed efficiency, but the inclusion of CFM in the experimental diets increased mortality, decreased weight gain and decreased feed efficiency. In both experiments, increased quantities of dietary cassava cyanate were associated with increased (P < 0.05) blood serum thiocyanate concentrations. Histopathological examination of thyroid, liver and kidney revealed no appreciable alterations due to the cassava feeding, however there was no conclusive evidence of cyanide or thiocyanate effects on thyroid activity. Aflatoxin contamination appeared to have contributed to the high mortality rate associated with CFM diets. The results showed that broilers were tolerant of relatively high levels of dietary cyanogenic glycosides (Gomez et al., 1988). 2.2.3

Long-term/carcinogenicity studies No information available.

2.2.4

Reproduction studies

2.2.4.1

Rats

In a one-generation reproduction study albino female rats (10 rats/group) were fed ad libitum one of the following diets: A, a normal laboratory diet, B, a 50% gari diet (Nigerian preparation of cassava), C, a raw cassava diet, D, a diet containing 5 g KCN/100 g and E, a diet containing 10 g KCN/100 g. After 2 weeks rats in each group were mated with 5 adult males fed normal diet. Pregnant rats from each group were maintained on their respective diets. After littering, the newborn rats were studied for postnatal development. After 21 days F1 rats were put for another 4 weeks on their respective diets. The offspring of the rats fed the 50% gari diet had significantly lower birth weights and brain weights and never attained the same adult weights as those of the controls. The adult female rats fed a diet consisting entirely of raw cassava had significantly reduced haematological and biochemical parameters (Hb, PCV, serum protein and T4 concentration). This diet also caused an increased incidence of cannibalism and a significant reduction in the frequency of pregnancy, the average number of pups per litter and birth weights among these pups. In addition there was an increased incidence of neonatal deaths among the offspring which also had poor development, reduced brain weights and an increased tendency of aggression towards their litter mates. Adult female rats fed diets containing 5 and 10 g KCN/100 g laboratory diet survived for more than three months but never became pregnant. They developed enlarged thyroid glands and tumours of the large intestine. The usual content of cyanide in cassava varies from 70 to 500 mg/kg which is much less than the levels used in these experiments; thus the rats were able to cope with the 50% gari diet and detoxify the glycoside present (Olusi et al., 1979). 2.2.4.2

Pigs

General toxicity and reproductive effects were studied for cassava in combination with added cyanide. In a 110-day feeding experiment 18 pregnant Yorkshire gilts were allocated to three equal groups and fed fresh cassava (containing 40.2 HCN/kg) supplemented with 0, 250, and 500 mg cyanide (KCN) per kg of fresh cassava offered. Serum thiocyanate concentration was slightly but not significantly increased in the 500 mg KCN/kg group and serum protein bound iodine decreased during gestation in all groups. Fetal serum thiocyanate concentration was significantly (p <0.05) higher in the group fed 500 mg KCN/kg. A small increase in maternal thyroid weight with increasing levels of cyanide was observed. Pathological studies showed proliferation of glomerular cells of the kidneys in gilts of all groups and reduced activity of the thyroid gland in gilts fed 500 mg KCN/kg group. Cyanide fed during gestation did not affect performance during lactation. Milk thiocyanate and colostrum iodine concentrations were significantly higher in the group fed 500 mg KCN/kg feed. No effects of cyanide were reported on indices of reproduction performance (Tewe & Maner, 1981a). 2.2.5

Special studies on embryotoxicity and/or teratogenicity

In a teratogenicity study pregnant hamsters received oral doses of 70, 100, 120 or 140 mg linamarin/kg bw or an equivalent volume (0.5 ml/100 g) of isotonic saline during the early primitive streak stage of gestation (day 8 of gestation). The hamsters were killed on the morning of day 15 of pregnancy. Fetuses were removed by caesarian section and the numbers of resorption sites, dead fetuses, and living fetuses were recorded. Living fetuses were examined for gross external malformations and by means of histopathological methods for internal malformations. A dose of 120 or 140 mg linamarin/kg bw was associated with an increased incidence of vertebral and rib anomalies as well as the production of encephaloceles in the offspring. These larger doses of linamarin also resulted in obvious maternal toxicity (dyspnoea, hyperpnoea, ataxia, tremors and hypothermia). Two animals dosed with 140 mg and one animal dosed 120 mg/kg bw died. In surviving animals the signs of poisoning were greatly reduced or gone within 3 h after treatment. Linamarin treatment had no effect on fetal body weight, ossification of skeletons, embryonic mortality, nor litter size. Although ingestion of the cyanogenic glycoside was associated with a significant teratogenic response, the effects occurred only at doses that elicited signs of maternal intoxication (Frakes et al., 1985). In a teratogenic study groups of pregnant hamsters (8 dams/group) were fed diets consisting of cassava meal:laboratory chow (80:20) during days 3-14 of gestation. One low cyanide (sweet) cassava meal and one high cyanide (bitter) cassava meal were studied. An additional group was fed a diet which resembled cassava in nutrtional value, but which lacked cyanogenic glycosides. Thiocyanate concentrations in the urine and blood of dams fed cassava diets increased significantly. Increased tissue thiocyanate concentrations were observed in fetuses recovered from cassava-fed dams. Cassava-fed dams gained significantly less weight than did control animals and their offspring showed evidence of fetotoxicity. Reduced fetal body weight and reduced ossification of sacrocaudal vertebrae, metatarsals and sternebrae were associated with cassava diets. High cyanide cassava diets were also associated with a significant increase in the numbers of runts compared to litters from dams fed either low protein or laboratory stock diets (Frakes et al., 1986). 2.2.6

Special studies on the thyroid gland

In a study cited by Oke (1980), the influence of a 100% cassava diet on the thyroid in a 7-day experiment with rats. A significant decrease in glandular stores of stable iodine, significantly higher thyroid weight and higher thyroidal 131I uptake were observed. Each effect is due to a synthetic block in the conversion of monoiodothyronine to diiodothyronine. 2.3 2.3.1

Observations in humans Acute toxicity studies in humans

One to 10 g of amygdalin have been given parenterally in humans, apparently without acute toxicity. This indirectly suggests that there is no significant metabolism of the intact injected glycoside. The cyanide-containing breakdown products possess well-defined toxicities, and 50 mg of hydrogen cyanide can be fatal. With oral dosing of amygdalin, a toxic potential is manifest. ß-Glucosidase is present in the gastrointestinal lumen, a contribution of intestinal microflora. According to Eyerly (1976) oral laetrile (amygdalin) could be 40 times more toxic than parenterally administered doses. This is probably due to the free HCN released by the ß-glucosidase enzyme present in the gut (Dorr & Paxinos, 1978). The lethal dose of amygdalin for man when ingested is reported to be in the range of 0.02-0.13 mmol/kg bw (Solomonson, 1981). If it is assumed that about 100-2000 mg HCN is the lethal dose for man, as much as 10-20 kg of Lafun cassava (10-20 mg cyanide/kg) will have to be consumed at a sitting to produce toxicity (Oke, 1980). Well-nourished individuals have ingested 1000 mg or more of pure amygdalin every day without any evidence of "side effects" (Oke, 1979). In a case study an 11-month-old girl was reported accidentally to have ingested 1-5 amygdalin tablets (500 mg). The patient became listless within one half hour of ingestion and vomited. Breathing became irregular and her state of consciousness became altered. An hour after ingestion she was in shock and died approximately 72 h following ingestion in spite of hospital treatment (Humbert et al., 1977). In a case-study a 17-year-old girl suffering from cancer made a practice of taking, instead of radiotherapy, four ampoules of laetrile (3 g amygdalin) intravenously. One day she swallowed three 1/2 ampoules of laetrile. Shortly after ingestion, a severe headache and dizziness developed, and she collapsed. Laboured breathing developed, her pupils became dilated, and she became comatose. All symptoms occurred within 8-10 minutes after ingestion. She died 24 h after ingestion (Sadoff et al., 1978). In Anatolia (Turkey) 9 cases of cyanide intoxication of children due to the ingestion of wild apricot seeds (217 mg HCN/100g) were reported. The victims had probably eaten more than 10 seeds. Also in studies of Jeanin et al. (1961) and Pijoun (1942) poisoning after consuming a relative large amount of peach seeds or bitter almonds are reported. Quantitative figures on cyanogenic glycoside or cyanide intake are not given (Sayre & Kaymakcalavu, 1964). The toxicity of cassava and cassava processing products was until recently assumed to be associated with free cyanide, this was 50-60 mg which constitutes a lethal dose for an adult man. The cyanogenic glycosides were at first thought to be of little consequence to mammals if cassava hydrolytic enzymes have been

inactivated. The possibility of hydrolysis during digestion, however, is also important (Cooke & Coursey, 1981). In a case-study a 67-year-old woman collapsed after ingestion of a slurry of 12 bitter almonds ground up and mixed with water. She recovered after treatment in the hospital. The average cyanide content was 6.2 mg HCN/bitter almond (Shragg et al., 1982). The consumption of 60 bitter almonds is deadly for an adult. For young children, however 5-10 almonds or 10 droplets of bitter almond oil are fatal (Askar & Moral, 1983). 2.3.2

Long-term toxicity studies in humans with cyanogenic glycosides and cyanides

A study to evaluate the possible association of high cyanide and low sulfur intake in cassava-induced spastic paraparesis was performed. The north-eastern part of Mozambique suffered a severe drought in 1981: the only crop available was the most toxic variety of cassava and, due to lack of food during the harvest period, the roots were eaten after only a few days of sun drying. A field survey revealed 1102 cases of spastic paraparesis. In 1982 urine was collected from 30 apparently healthy children (age 8.1 years). As reference 17 Swedish children (age 8.6 years) were used. In a second stage urine was sampled in 1983 (when the nutritional situation was improved but still unsatisfactory) from 31 children (9.0 years) in the same village and 30 schoolchildren (8.1 years) in a nearby district where no cases of paraparesis were seen in 1981 and from 28 children (7.1 years) of the city who ate virtually no cassava. The children from the village had increased thiocyanate and decreased inorganic sulfate excretion, indicating high cyanide and low sulfur-containing amino acid intake. Children from a neighbouring cassava-eating area, where no cases of spastic paraparesis had occurred, had lower thiocyanate excretion but higher organic sulfate excretion. These results support the hypothesis that the epidemic was due to the combined effects of high dietary cyanide exposure and sulfur deficiency (Cliff et al., 1985). Several studies were performed, including epidemiological studies, on the role of chronic cyanide intoxication caused by the consumption of cassava diet in the etiology of tropical (ataxia) neuropathy (TAN) in Nigerian populations. As described in over 400 Nigerian patients the essential neurological components of the disease are myelopathy, bilateral optic atrophy, bilateral perceptive deafness and polyneuropathy. The initial and most common symptoms consist of various forms of paraesthesia and dysaesthesia usually starting in the distal part of the lower limbs. The next most common finding is blurring or loss of vision. Other common symptoms in order of frequency are ataxia, tinnitus, deafness, weakness and thinning of the legs. In about a third of the patients stomatoglossitis is present, additionally motor neurone disease, Parkinson's disease, cerebellar degeneration, psychosis and dementia have been associated with the disease. TAN affects males and females in all age groups equally, but occurs only rarely in children under 10 years. Patients usually give histories of almost total dependence on a monotonous diet of

cassava derivatives; occasional dietary supplements include yam, maize, rice, vegetables and animal protein. Analysis of family relationships among patients showed no evidence of a genetically determined predisposition. The families were usually poor and members lived communally. Clinical evidence of malnutrition was frequently absent. The occurrence of cyanide intoxication in Nigerian patients was indicated by the significantly higher cyanide and thiocyanate plasma levels and higher excretion of thiocyanate than controls. Hepatic rhodanese activity was not different from that in controls and histology of liver biopsy specimens showed no abnormality. Total plasma vitamin B12 levels are normal or high in patients and healthy Nigerians but plasma concentration of cyanocobalamin was highly significantly raised in patients. A small proportion of cyanocobalamin was found in the liver of patients. Methylmalonic acid excretion was normal in patients, indicating that there was physiological vitamin B12 adequacy at tissue or cellular level in these patients. In Nigeria, endemic foci of the disease (in epidemiological studies) recognized since the early 1930's, correspond with the areas where cassava is intensively cultivated and consumed as the major or the sole dietary source of carbohydrate. There was evidence of increased exposure to cyanide in members of families where multiple cases were found. In biochemical studies no biochemical evidence of protein-calorie malnutrition was seen and serum transferrin, said to be sensitive index of protein nutritional status, was normal. No haematological abnormalities were seen in patients. Investigation of patient nutritional status with regard to the water-soluble vitamins showed no abnormality, except in riboflavin intake. Plasma concentrations of calcium, phosphate, sodium, potassium, chloride, bicarbonate, and cholesterol, and tests of thyroid, hepatic, and renal functions were normal. Amino acids and porphyrobilinogen and other porphyrins were absent in urine. There was no biochemical nor other evidence of malabsorption. Glucose tolerance tests were normal. Histamin-fast achlorhydria was very rare. Serologic tests for syphilis, typhoid, typhus, brucellosis and screening for prevalent viral infections gave negative or insignificant results. There was no increased prevalence of malaria and urinary tract infections were not encountered in patients. ECG's were normal. Diminished urinary excretion and riboflavin (vitamin B2) and low serum riboflavin and caeruloplasmin levels in patients compared with healthy persons were the only significant abnormalities found. Riboflavin deficiency is, however, widespread in many parts of Nigeria and especially in areas endemic for TAN, probably because cassava is a poor source of the vitamin. The daily intake of HCN from cassava derivatives in areas in Nigeria endemic for TAN may be as high as 50 mg, which is nearly an acutely sublethal amount, and can conceivably produce cyanide intoxication. This is particularly plausible as cassava root is a major item of food, often the main source of carbohydrate, and is poor in sulfur-containing amino acids which are essential for detoxification of cyanide. A high prevalence (2%) of goitre in populations with a high incidence of TAN is seen; this appears to be related to cassava

diet and high plasma thiocyanate. The effects of riboflavin insufficiency may combine with those of chronic cyanide intoxication in the etiology of TAN (Osuntokun, 1981). Several epidemiological and experimental studies revealed that TAN resulted from chronic cyanide poisoning due to the release of HCN from cyanogenic glycosides present in certain cassava food products. Since TAN invariably occurs among poorly nourished people, the condition may result from unidentified nutritional deficiencies or excesses as well as cyanide toxicity. Among patients showing the TAN syndrome the frequency of goitre was increased (Conn, 1979a,b). From both experimental and epidemiological studies there is strong, but not conclusive, evidence that cassava toxicity is a causative factor in some neurological disorders like TAN and endemic spastic paraparesis. As well, different deficiencies such as deficiency of sulfur-containing amino acids may play a role. The geographical distribution of malnutrition-related diabetes coincides with that of cassava consumption. Dietary exposure to cyanide has been proposed as a possible cause, but has not been proven. The poor nutritional quality of cassava seems a very likely causative factor (Rosling, 1987). Workers in certain occupations are exposed to HCN additional to sources encountered by the general public. These individuals have been studied to determine whether chronic cyanide poisoning is a clinical entity. Symptoms reported were headache, vertigo, tinnitus, nausea, vomiting, and tremors. Although these symptoms are sufficiently documented and characteristic, they are transitory in that exposure to fresh air causes recovery. These do not seem to produce the outward symptoms of TAN, the pathological condition that has been attributed in part to exposure to cyanide or cyanogenic glycosides in certain preparations of cassava (Conn, 1979b). Neurological disorders such as ataxic neuropathy and cretinism have been associated directly with the intake of cyanogenic glycoside-contaminated diets. There are grounds to suspect that cyanogenic glycoside-contaminated foodstuffs such as cassava and pulses are directly implicated in acute and chronic cyanide toxicity in the tropics. Although correlation between dietary cyanogen-contaminants and disease such as TAN, goitre, cretinism and mental retardation exists and is based on experimental evidence, the mechanism of action of cyanide and thiocyanate on the cellular level is not well understood (Nartey, 1980). Based on several studies in humans consuming as their main food cassava or other food products rich in cyanogenic glycosides, it seems that chronic cyanide intoxication in combination with deficient intake of riboflavin and/or a poor quality of protein and hence methionine deficiency is/are responsible, to a large extent, for the etiology of TAN in cassava-eating areas, whereas chronic cyanide intoxication in combination with a deficient iodine intake is responsible for goitre (Oke, 1979, 1980). Epidemiological and experimental studies show that cyanogenic glycosides in food products play a important role in the development of goitre. Thiocyanate, the detoxification product from the HCN

derived from cyanogenic products, is responsible for interference with thyroid function. Studies on endemic goitre in Africa have identified iodine deficiency and an antithyroid activity of cassava diets as major etiological factors of the disease (Conn, 1979a,b). Extensive studies in Zaire have established that goitre and cretinism due to iodine deficiency can be considerably aggravated by a continuous dietary cyanide exposure from insufficiently processed cassava. This effect is caused by thiocyanate. Thiocyanate has a similar size to the iodine molecule and interferes with the iodine uptake in the thyroid gland. Thiocyanate levels which can occur after exposure to cyanide from cassava can only affect the gland when the iodine intake is below 100 micrograms/day, which is regarded minimal for normal function. Several studies have shown that populations with considerable cyanide exposure from inadequately processed cassava are free from goitre as long as their iodine intake is sufficient. Populations in northern Zaire with very low iodine intake and in addition high thiocynate levels resulting from the consumption of cassava products, show very severe endemic goitre. When the population was given iodine supplementation, the goitre decreased (reviewed by Rosling, 1987). In persons who ingested cassava a decreased 131I-uptake by the thyroids was seen, confirming the goitrogenous nature of cassava (De Lange, 1973 as cited by Oke, 1980). In one study more than 50% of workers in a processing factory revealed thyromegaly and an increased 131I-uptake. All workers had increased blood haemoglobin concentration as well as increased lymphocyte counts. Two of the employees developed psychosis, although it is difficult to know whether this was cause or effect. Abnormal thyroid function has also been found in workers from a photographic plant in which a cyanide extracting process was used to recover silver from X-ray films (El Ghawabi et al., 1975). Alterations in vitamin B12 and folate metabolism have also been noted to play a role, although the significance of these observations are not known. Short of having an elevated whole blood cyanide level, the laboratory findings of cyanide poisoning are fairly non-specific. Perhaps the most valuable clue is the finding of a severe metabolic acidosis with a high anion gap. In clinical medicine there have been reported only 7 causes of anion gap metabolic acidosis. In most of these disorders lactic acidosis is present owing to severe tissue underperfusing and hypoxia. In the patient with cyanide poisoning tissue hypoxia is universally present (Gonzales & Sabatini, 1989). 3.

COMMENTS AND EVALUATION

The potential toxicity of a cyanogenic plant depends primarily on its capacity to produce a concentration of hydrogen cyanide toxic to animals and humans. The release of hydrogen cyanide can occur either following maceration of the plant material - this activates the intracellular ß-glucosidase which in turn hydrolyses glycoside or by hydrolysis of glycoside by the ß-glucosidase produced by the microflora of the gut. The level of ß-glucosidase activity in the

gut depends on the pH and the bacterial composition. The cyanogenic glycoside content of a foodstuff, when known, is usually expressed in terms of the amount of cyanide released by acid hydrolysis; exact figures for the concentration of the glycosides themselves are very rarely given. Hydrogen cyanide absorbed from the gut can be detoxified by metabolic conversion to thiocyanate; this depends on the presence of nutritional factors, such as sulfur-containing amino acids and vitamin B12. Acute toxicity results when the rate of absorption of hydrogen cyanide is such that the metabolic detoxification capacity of the body is exceeded. Available reports of toxicological studies lack information on the level of intake of cyanogenic glycosides or on the amount of hydrogen cyanide potentially released. No long-term toxicity or carcinogenicity studies were available to the Committee. However, in vitro and in vivo genotoxicity were negative. Teratogenic and adverse reproductive effects attributable to linamarin (cassava) and hydrogen cyanide were seen only at doses that also caused maternal toxicity. The toxic effects of cyanide on the thyroid (via its metabolite thiocyanate) depend on the iodine status of the test animals, as indicated in the twenty-fifth report of the Committee (Annex 1, reference 56). On the basis of epidemiological observations, associations have been made between chronic exposure to cyanogenic glycosides and diseases such as spastic paraparesis, tropical ataxic neuopathy, and goitre. However, these observations were confounded by nutritional deficiencies, and causal relationships have not been definitely established. Traditional users of foods containing cyanogenic glycosides usually have a basic understanding of the treatment required to render them safe for consumption. However, some products are sold commercially and are consumed by people who may not be familiar with such procedures. The Committee therefore recommended that guidelines be developed to provide reliable and sensitive methods for the analysis of these foodstuffs for hydrogen cyanide releasable from cyanogenic glycosides, in order to ensure that amounts in foods as consumed do not present a hazard. Because of a lack of quantitative toxicological and epidemiological information, a safe level of intake of cyanogenic glycosides could not be estimated. However, the Committee concluded that a level of up to 10 mg/kg hydrogen cyanide in the Codex Standard for Cassava Flour (CAC, 1991) is not associated with acute toxicity. 4.

REFERENCES

ASKAR, A., & MORAD, M.M. (1983). Lebensmittelvergiftigung 1. Toxine in natürlichen Lebensmittel. Alimentia, 19: 59-66.

BARRETT, M.D., HILL, D.C., ALEXANDER, J.C. & ZITNAK, A. (1977). Fate of orally dosed linamarin in the rat. Can. J. Physiol. Pharmacol., 55: 134- 136. BASS, N.H. (1968). Pathogenesis of myelin lesions in experimental cyanide poisoning: a microchemical study. Neurology, 18: 167-177, as cited in Osuntokun, 1981. BASU, T.K. (1983). High-dose ascorbic acid decreases detoxification of cyanide derived from amygdalin (laetrile): Studies in guinea-pigs. Can. J. Physiol. Pharmacol., 61: 1426-1430. BRIERLEY, J.B., PRIOR, P.F., CALVERLEY, J. & BROWN, A.W. (1977). Cyanide intoxication in Macaca mulatta: Physiological and neurological aspects. J. Neurol. Sci., 31: 133-157, as cited in Osuntokun, 1981. CAC (1991). Codex Alimentarius, Vol. XII, Suppl 4. Codex Standard for Edible Cassava Flour (African Regional Standard). Rome, Food and Agriculture Organization of the United Nations (CODEX STAN 176). CLARK, A. (1936). Report on effects of certain poisons contained in food plants of West Africa upon health of native races. J. Trop. Med. Hyg., 39: 285-295, as cited in Osuntokun, 1981. CLIFF, I., LUNDQUIST, P., MÄRTENSSON, I., ROSLING, H. & SÖRBO, B. (1985). Association of high cyanide and low sulfur intake in cassava-induced spastic paraparesis. Lancet, II: 1211-1213 COLLIER JACKSON, J. (1988). Behavioral effects of chronic sublethal dietary cyanide in an animal model: implications for humans consuming cassava (Manihot esculenta). Journal of the Society for the Study of Human Biology, 60: 597-614. CONN, E.E. (1979a). Cyanide and cyanogenic glycosides. In Rosenthal, G.A. & Janzen, D.H. (eds.), Herbivores: Their interaction with secondary plant metabolites, Academic Press, Inc., New York-London, pp 387-412. CONN, E.E. (1979b). Cyanogenic glycosides. International review of biochemistry. In Biochemistry and Nutrition 1A, Neuberger, A., & Jukes, T.H. (eds), University Park Press, Baltimore, 27: 21-43. COOKE, R.D. & COURSEY, D.G. (1981). Cassava: A major cyanide-containing food crop. In Vennesland, B., Conn, E.E., Knowles, W, Westley, J. & Wissing, F. (Eds). Cyanide in Biology. Academic Press, London. DOHERTY, P.A., FERN, V.H. & SMITH, R.P. (1982). Congenital malformations induced by infusion of sodium cyanide in the Golden hamster. Toxicology and Applied Pharmacology, 64: 456-464. DORR, R.T. & PAXINOS, I. (1978). The current status of laetrile. Annals of Internal Medicine, 89: 389-397. EPA (1990). Summary Review of Health Effects Associated with Hydrogen Cyanide, Health Issue Assessment Environmental Criteria and

Assessment Office, Office of Health and Environmental Assessment Office of Research and Development, US Environmental Protection Agency Research Triangle Park, North Carolina, USA. EYERLY, R.C. (1976). Laetrik; focus on the the facts. 50-54.

Cancer, 26:

FERRARO, A. (1933). Experimental toxic encephalomyelopathy (diffuse sclerosis following subcutaneous injection of postassium cyanide). Archs. Neurol. Psychiat., 29: 1364-1367, as cited in Osuntokun, 1981. FRAKES, R.A., SHARMA, R.P. & WILLHITE, C.C. (1985). Development toxicity of the cyanogenic glycoside linamarin in the golden hamster. Teratology, 31: 241- 246. FRAKES, R.A., SHARMA, R.P., WILLHITE, C.C. & GOMEZ, G. (1986). Effect of cyanogenic glycosides and protein content in cassava diets on hamster prenatal development. Fundamental and Applied Toxicology, 7: 191-198. FRAKES, R.A., SHARMA, R.P. & WILLHITE, C.C. (1986). Comparative metabolism of linamarin and amygdalin in hamsters. Food Chemical Toxicology, 24: 417-420. GEITLER, A.O. & BAINE, J.O. (1983). The toxicity of cyanide. Med. Sci., 195: 182-198.

Am. J.

GOMEZ, G., APARICIO, M.A. & WILLHITE, C.C. (1988). Relationship between dietary cassava cyanide levels and Brailer performance. Nutrition Reports International, 37: 63-75. GONZALES, I. & SABATINI, S. (1989). Cyanide poisoning: pathophysiology and current approaches to therapy. The Int. I. of Artificial Organs, 12(6): 347-355. HIMWICH, W.A. & SAUNDERS, I.P. (1984). Enzymatic conversion of cyanide to thiocyanate. American Journal of Physiology, 58: 348. HUMBERT, I.R., TRESS, I.H. & BRAICO, K.T. (1977). Fatal cyanide poisoning: accidental ingestion of amygdalin. JAMA, 238: 482. HURST, E.W. (1940). Experimental demyelination of the central nervous system. Aust. J. Exp. Biol. Med. Sci., 18: 201-223, as cited in Osuntokun, 1981. LANG, K. (1933) Die Rhodanbildung im Tierkörper. 243-256.

Biochem Z., 259:

LESSELL, S. (1971) Experimental cyanide optic neuropathy. Ophtal., 84: 194-204, as cited in Osuntokun, 1981.

Archs.

LEUSCHNER, F., NEUMANN, B.W., & LIEBSCH, M. (1983a). Mutagenicity study of hydrocyanic acid in the Ames Salmonella/microsome plate test (in vitro). Unpublished study, Laboratory of Pharmacology and Toxicology, Hamburg, August 1983, submitted to the WHO by Detia Freyberg GmbH.

LEUSCHNER, F., NEUMANN, B.W., & LIEBSCH, M. (1983b). Mutagenicity study of hydrocyanic acid in Chinese hamster (chromosome aberration) by oral administration. Unpublished study, Laboratory of Pharmacology and Toxicology, Hamburg, August 1983, submitted by Detia Freyberg GmbH. LEUSCHNER, F. & NEUMANN, B.W. (1989a). In vitro mutation assay of KCN in Chinese hamster cells. Unpublished study, Laboratory of Pharmacology and Toxicology, July 1989, submitted by Detia Freyberg GmbH. LEUSCHNER, F., NEUMANN, B.W., OTTO, H. & MÖLLER, E. (1989b). 13-Week toxicity study of potassium cyanide administered to Sprague-Dawley rats in the drinking water. Unpublished study, Laboratory of Pharmacology and Toxicology, July 1989, submitted by Detia Freyberg GmbH. MAJAK, W., MCDIARMID, R.E., JAKOBER, K. & CHENG, K.I. (1989). Diurnal changes in rates of degradation of cyanogenic glycosides in bovine rumen fluid. Toxicon., 27: 61. NARTEY, F. (1980). Toxicological aspects of cyanogenesis in tropical foodstuffs in Toxicology in the Tropics. Editors R.L. Smith and E.A. Bababumni, Taylor & Francis Ltd, London, 53-73. OKE, O.L. (1979). Some aspects of the role of cyanogenic glycosides in nutrition. Wld. Rev. Nutr. Diet, 33: 70-103. OKE, O.L. (1980). Toxicity of cyanogenic glycosides. Chemistry, 6: 97-109.

Food

OKOH, P.N. & PITT, G.A.J. (1981). The metabolism of cyanide and the gastrointestinal circulation of the resulting thiocyanate under conditions of chronic cyanide intake in the rat. Can. J. Physiol. Pharmacol., 60: 381-385. OKOH, P.N. (1983). Excretion of 14C-labeled cyanide in rats exposed to chronic intake of potassium cyanide. Toxicology and Applied Pharmacology, 70: 335-339. OLUSI, S.O., OKE, O.L. & ODUSOTE, A.C. (1979). Effects of cyanogenic agents on reproduction and neonatal development in rats. Biol. Neonate, 36: 233-293. OSUNTOKUN, B.O. (1981). Cassava diet, chronic cyanide intoxication and neuropathy in the Nigerian Africans. Wld. Rev. Nutr. Diet, 36: 141-173. PHILBRICK, D.I., HOPKINS, I.B., HILL, D.C., ALEXANDER, I.C. & THOMSON, R.G. (1979). Effects of prolonged cyanide and thiocyanate feeding in rats. Journal of Toxicology and Environmental Health, 5: 579-592. RAUWS, A.G., OLLING, M. & TIMMERMAN, A. (1982). The pharmacokinetics of amygdalin. Arch. Toxicol., 49: 311-319.

RAUWS, A.G., OLLING, M. & TIMMERMAN (1983). The pharmacokinetics of prunasin, a metabolite of amygdalin. J. Toxicol. Clin. Toxicol., 19: 851-856. ROSLING, H. (1987). Cassava toxicity and food security. Ed. Rosling. Tryck Kontakt, Uppsala, Sweden, 3-40. SADOFF, L., FUCHS, K. & HOLLANDER, I. (1978). Rapid death associated with laetrile ingestion. JAMA, 239: 1532. SAYRE, I.W. & KAYMAKCALAVU, S. (1964). Cyanide poisoning from apricot seeds among children in Central Turkey. New. Engl. J. Med., 270: 1113-1118. SHRAGG, T.A., ALBERTSON, T.E. & FISHER, C.J. (1982). Cyanide poisoning after bitter almond ingestion. The Western Journal of Medicine, 136: 65-69. SMITH, A.D.M., DUCKETT, S. & WATERS, A.H. (1963) Neuropathological changes in chronic cyanide intoxication. Nature, 200: 179-181, as cited in Osuntokun, 1981. SOLOMONSON, L.P. (1981). Cyanide as a metabolic inhibitor. In Cyanide in Biology, Vennesland, B., Conn, E.E., Knowles, C.J., Westley, J & Wissing, F. Academic Press, London, New York, Toronto, 11-18. TEWE, O.O. & MANER, I.H. (1980). Cyanide, protein and iodine interactions in the performance, metabolism and pathology of pigs. Research in Veterinary Science, 29: 271-276. TEWE, O.O. & MANER, I.H. (1981a). Performance and patholophysiological changes in pregnant pigs fed cassava diets containing different levels of cyanide. Research in Veterinary Science, 30: 147-151. TEWE, O.O. & MANER, I.H. (1981b). Long-term and carry-over effect of dietary inorganic cyanide (KCN) in the life cycle performance and metabolism of rats. Toxicological and Applied Pharmacology, 58: 1-7. UMOH, L.B., MADUAGWA, E.N. & AMOLE, A.A. (1986). Fate of ingested linamarin in malnourished rats. Food Chemistry, 20: 1-9. VENNESLAND, B., CASTRIC, P.A., CONN, E.E., SOLOMONSON, L.P., VOLINI, M. & WESTLEY, I. (1982). Cyanide metabolism. Fed. Proc., 41(10): 2639-2648. WILLHITE, C.C. (1982). Congenital malformations induced by laetrile. Science, 215: 1513-1515. WILLIAMS, A.O. & OSUNTOKUN, B.O. (1969). Light and electron microscopy of peripheral nerves in tropical ataxic neuropathy. Archs. Neurol., 21: 475-492, as cited in Osuntokun, 1981. WHO (1965). Evaluation of the hazards to consumers resulting from the use of fumigants in the protection of food. Report of the second

Joint Meeting of FAO Committee on Pesticides in Agriculture and the WHO Expert Committee on Pesticides Residues, FAO Meeting Report No Pl/1965/10/2: WHO Food series 28.65. See Also: Toxicological Abbreviations CYANOGENIC GLYCOSIDES (JECFA Evaluation)