LD50 (oral, rat) bitter almonds

LD50 (oral, rat) bitter almonds is estimated by kinetic computer models based on literature values. The model incorporates the hydrolysis of amygdalin by the emulsin complex to HCN in the mouth, the processing of amygdalin by the body to HCN and excretion in the urine, the blood elimination of HCN to thiocyanate.

 

 

SATOCONOR.COM

J.G. van der Galiën ‘Estimation Of The LD50 (Oral, Rat) Bitter Almonds By Kinetic Computer Models’ 4.3. (2005)

Full paper

SATOCONOR.COM Journal of Toxicology

 

 

Estimation Of The LD50 (Oral, Rat) Bitter Almonds By Kinetic Computer Models

How save is the alternative anti-cancer cure?

By Johan G. van der Galiën (M.Sc.)

(For comments e-mail: johan.van.der.galien@satoconor.com)

Version 1.1. April 6, 2006 (version 1.0. from August 27, 2005)

 

HOME of SATOCONOR.COM

 

Abstract:

The literature on the Internet and scientific journals is reviewed about the metabolism of amygdalin from bitter almonds by mammals. Also the chemistry of the base and acid catalysed hydrolysis of the aglucone mandelonitrile from amygdalin is discussed, because not all reactions are biochemical. The pH dependent stability of acetone cyanohydrin / water mixtures are used as a model for what happens with mandelonitrile during chemical hydrolysis in the pH range 6.0 – 7.4. This is the range at which the other parts of the mammal body operate, beside the stomach (pH = 1 – 3). The chemical hydrolysis in this range is the formation of a carbonyl compound and HCN. The bonded mandelonitrile moiety in amygdalin or in the metabolic intermediate prunasin will not give this reaction.

The fact that about 8% of the bonded HCN in bitter almonds is released in the mouth, with an average chewing time of one minute, is made plausible. T1/2 amygdalin in the mouth, with beta-glucosidases and cyanohydrin lyase present, is estimated to be about 8.3 minutes. Also the fact that there is no release of HCN in the strong acidic environment of the stomach is discussed. Most likely free en bonded mandelonitrile is then hydrolysed to some extent to the beta-hydroxyamide and from there to mandelic acid.

The LD50 (oral, rat) bitter almonds is roughly estimated by two computer models, that take some biochemical and chemical reactions in the mouth and body in to account, to be 3.3 g / Kg body weight. This number is derived by the most advanced model from a LD50 (oral, rat) HCN = 0.81 mg / Kg body weight, T1/2 amygdalin in the mouth = 8.3 minutes (from experiments in water with the emulsin complex of almond pulp), T1/2 amygdalin in the body = 806 minutes (calculated with a simpler computer model), T1/2 excretion amygdalin in urine = 13,000 minutes (calculated from literature data of rats) and T1/2 HCN elimination form the blood = 14.1 minutes (literature data of rats, regarded as the processing rate by the whole body). Care must be taken by extrapolating this LD50 from the computer model to other mammals, because the biochemistry involved can differ dramatically. Despite the fact that the LD50 (oral, rat) HCN is among the lowest known to mammals, There is a contradiction, because it is also said in the literature that cyanide is less toxic for rats, then for dogs, rhesus monkeys and rabbits, because rats have a faster rhodanese based detoxifying system.

The computer model gives a time of death of 61 seconds when rats eat 3.3 g bitter almonds per Kg body weight, indicating that the 60 seconds chewing time in the mouth, which is fixed in the model, releases most of the deadly dose (99.98%). Only 0.02% is released in the body. Practically and within the limitations of the computer model one can say that LD50 (oral, rat) bitter almonds is totally determined by the biochemical action of the emulsin complex on amygdalin and the chemical hydrolysis of mandelonitrile during chewing time.

According to the estimations presented in this paper bitter almonds are about as toxic as ordinary kitchen salt. LD50(oral, rat) ≈ 3 gram / Kilogram body weight. But I also learned from the Toxicology newsgroup that estimations and predictive methods are no substitute for the values found by experiment.

 

1. Introduction

1.1 The nitrilosides

Amygdalin (1.) is one of the 14 known natural occurring nitrilosides. Some of the others are called prunasin, dhurrin, and linamarin. All the different molecules have a sugar moiety and a non-sugar moiety (aglucone) comprised of hydrogen cyanide (HCN) and benzaldehyde (Ph-CH=O) or para-hydroxybenzaldehyde (p-HO-Ph-CH=O) or acetone (CH3-C(=O)-CH3) molecule.

Sugar-O-CH(Ph)-CN (Amygdalin, sugar is the disaccharide gentiobiose)

1.

Or

Sugar-O-CH(Ph)-CN (Prunasin, sugar is the monosaccharide glucose)

2.

Or

Sugar-O-CH(Ph-OH-p)-CN (Dhurrin, sugar is the monosaccharide glucose)

Or

Sugar-O-C(CH3)2-CN (Linamarin, sugar is the monosaccharide glucose)

Fig. 1: 5 of the 14 known nitrilosides.1 Ph = -C6H5

There are over 1200 plant species that contain nitrilosides. They come from all over the world, including from arctic regions. Some of these plants are edible for animals and humans. O.L. Oke has noted that cyanogenetic glycosides (= nitrilosides) have been found in the following common vegetables: maize, sorghum, millet, field bean, lima bean, kidney bean, sweet potato, cassava, lettuce, linseed, almond and seeds of lemons, limes, cherries, apples, apricots, prunes, plums and pears.2

I will focus on amygdalin from bitter almonds, the tree on which they grow is called Prunus amygdala, in this paper. But I must state that in all the mentioned plants and seeds the amygdalin is chemical identical with that in bitter almonds. It can be that they all do not have the emulsin complex of enzymes to degrade the amygdalin to HCN, benzaldehyde and glucose, which certainly can be found in bitter almonds.

The nitrilosides are interesting because they are intended to function as a defence mechanism of the plant in question against attack by herbivores, insects and micro-organisms. When an organisms tries to feed on the plant the nitrilosides and HCN realising enzymes, which are stable stored in different cell compartments, come together releasing the potentially deadly HCN which can kill the organism when it digests enough of the plant in question.

Most likely this was also a defence mechanism against herbivores in the early stages of evolution. Today sheep can digest pasture grasses, which contain about 3% amygdalin in the dry state1; their whole live long without any harm! And the daily intake is 2.7 Kg (dry matter) per ewe.3 Since the 3% is also based on dry matter this corresponds to a daily intake of 81 grams of amygdalin. They graze 9.7 hours a day. It is also said by Krebs Jr that 95% of the amygdalin is converted to HCN by the sheep within one hour.1 My estimation of the T1/2 (amygdalin, sheep) is 14 minutes.4 So this brings us to about 81 * 27 / 457 = 4.9 g of pure HCN released in the sheep body in one grazing day. (LD50 (oral, sheep) amygdalin > 81 * 1000 / (40 * 9.7) = 210 mg / Kg body weight / hour.) Since already 95% is released in one hour The weight of an adult female sheep (ewe) is 40 Kg.6 The dose HCN = 4.9 * 1000 / 40 * 9.7 = 13 mg / Kg body weight / hour. So there must be a biochemical route to detoxify the HCN in sheep very fast because the lethal dose for HCN (oral, rat) when 50% of the poor rats are killed (LD50) is only 0.81 mg / Kg body weight.7 So herbivores like sheep have adapted to the amygdalin / HCN defence of plants.

The full chemical analysis of bitter almonds, which is the natural resource with the highest amygdalin content known to man, can be found on the internet.8 Although this is very good reference the amygdalin content is not mentioned specifically, it falls under carbohydrate content. Other literature sources give 2 – 4%9, 4.2%10 and 4.9 – 5.2%7 for the amygdalin content of bitter almonds. I use the mean of7 5.1% in the rest of the publication.

 

1.2. Chemistry of the aglucone moiety of amygdalin (mandelonitrile)

The non-sugar part of amygdalin is called mandelonitrile (3.).

 

HO-CH(Ph)-CN

3.

Fig. 2: Mandelonitrile

 

It belongs to the so-called cyanohydrin family. Cyanohydrins have as main chemical formula R-COH(CN)-R’ whereby -COH(CN)- is the cyanohydrin group. Mandelonitrile is comprised by the –OH, hydroxyl or alcohol group, the –CN, cyano or nitrile group and R = –Ph, a benzene ring or phenyl group, and R’ = H. As you can see here there are many synonyms in Organic Chemistry. In synthesis cyanohydrins are seldom isolated as the pure compound, because most of them are unstable. Most of the times the raw product is directly used as a precursor for reactions to more stable compounds. The reaction mechanism for the formation of cyanohydrins by chemical methods is given in Fig. 3.

 

HCN + B ß à CN- + BH+

R-CO-R’ß CN- à R-CO-(CN)-R’ ß BH+ à R-COH(CN)-R’

 

Fig. 3: The reaction mechanism for the formation of cyanohydrins. R can be an alkyl or aryl group. R’ can be alkyl or aryl then the product is a cyanohydrin from a ketone. Or R’ can be hydrogen (H) then the product is a cyanohydrin from an aldehyde. B is the catalyst and stands for (strong) base. So the reaction is base catalysed.11

 

The reaction in Fig. 3 is reversible through the equilibria. By adding water (pH = 7, neutral conditions) to a pure cyanohydrin some of the bonded HCN is released because of the small concentration of OH- ions in pure water. These hydroxide ions can abstract a proton from the hydroxyl group of the cyanohydrin and in time equilibrium will be established. By adding high concentrations of a strong base you get almost all ketone or aldehyde back, because then the equilibria of Fig. 3 are shifted to the left. How is this possible? Is this a contradiction with the fact that you need basic conditions to synthesise cyanohydrins? No, during synthesis you add a large amount of cyanide ions, which shifts the equilibria of Fig. 3 to the right. So these are other conditions then hydrolysis. But what about strong acidic hydrolysis conditions? Well then something completely different happens! Laymen think that this also releases HCN from the cyanohydrin. Now there are only two possible mechanisms:

·         The hydroxyl group gets protonated. And this gives a water molecule as leaving group leading to an energetically unfavourable carbocation next to a very strong electron withdrawing cyano group.11 So this will virtually not happen.

·         The cyano group gets protonated. Then the well-known hydrolysis reaction of normal nitriles (= not alpha-substituted with an hydroxyl group) can happen. The mechanism of this is shown in Fig. 4.

 

3. ß H+Cl-/H2O à HO-CH(Ph)-C+=NH ß à HO-CH(Ph)-C(H2O+)=NHß à HO-CH(Ph)-COH=NH ß à HO-CH(Ph)-CO-NH- ß H+ à

HO-CH(Ph)-CO-NH2 (4.)(Beta-hydroxyamide under the right circumstances these compounds can be isolated)ß H+ à HO-CH(Ph)-C+OH-NH2 ß H2O à HO-CH(Ph)-C(H2O+)OH-NH2

ß à HO-CH(Ph)-C(OH)2-NH3+ ß à HO-CH(Ph)-C+(OH)2 + NH3 ß à HO-CH(Ph)-COOH (5.) + NH4+ (+ Cl-) (Mandelic acid and ammonium chloride)

 

Fig. 4: The acid catalysed hydrolysis of mandelonitrile.12

 

As a matter a fact the acid hydrolysis of cyanohydrins is the best method of producing beta-hydroxy carboxylic acids. It even goes with retention of chirality. This means that there is no racemisation of an optical cyanohydrin!12 These kinds of reactions are typically done with one part concentrated HCl (molarity = 12) and one part water. So the pH = - log (12 / 2) = -0.8. Since HCL pKa = -7 I think that in this highly concentrated environment only one proton out of one million protons is bounded in a HCl molecule. The rest of the H+ is free, actually as H3O+. So the pH is as calculated. I postulate that the reaction of Fig. 3 also goes to some extent in the strong acidic environment of the stomach (pH = 1 – 313) with bonded and free mandelonitrile during digesting amygdalin from plants and seeds.

But under basic conditions the formation of free cyanide ions from mandelonitrile is not the only reaction. Some formation of the beta-hydroxy amide is most likely a side reaction. And this compound can in theory also react further to mandelic acid, just like is the case for acid catalysed hydrolysis. See Fig. 5.

 

Main reaction: 3. ß OH- à -O-CH(Ph)-CN ß à O=CH(Ph) + CN-

Side reaction 1: 3. ß OH- à HO-CH(Ph)-C(OH)=N- ß -H2O à

HO-CH(Ph)-C(OH)=NH + OH- ß à 4.

Side reaction 2: 4. ß OH- à HO-CH(Ph)-CO-(OH)-NH2 ß à 5. + -NH2

 

Fig. 5: The base catalysed hydrolysis of mandelonitrile.11

 

The fact that mandelonitrile will also hydrolyse to a beta-hydroxyamide by base catalysis is supported by the fact that, at a pH = 9 – 11, reaction time 1 – 8 hours and temperature of 80oC, several HO-CH(R)-CN’s are hydrolysed by base catalysis of Borax in 67% - 86% yield to the corresponding beta-hydroxyamide. Some times these yields where obtained by adding KCN, which shifts the main reaction of Fig. 4.14 I suppose that the boron atoms of Borax form a complex with the hydroxy group, so it is stabilised, and will not react to a carbonyl group, releasing HCN. In other words the –CN group is mainly available for base catalysed hydrolysis and will not split of. Also it is known that normal nitriles hydrolyse under both acidic as basic conditions to an amide, which can be isolated, or can react further to an carboxylic acid.11

On the Internet one can also find data of the stability of a cyanohydrin in water at various pH. This is as a matter a fact pure acetone cyanohydrin [= (CH3)2C(OH)CN], which also occurs bonded in a nitriloside called Linamarin from cassava. I think the data T1/2 = 57, 28 and 8 minutes for stability in water at pH = 4.9, 6.3 and 6.8 also give an idea how the stability is for mandelonitrile.15 One can extrapolate the data, and there is a T1/2 = 0 for pH >= 7. So according to this the free mandelonitrile will virtual dissociate at once at pH > 7. You can find other measurements for this system: 54.7, 31.2, 5.4, and 4.0 min at pH 6.00, 6.40, 6.86, and 7.00.16 These show that the curve goes asymptotically to zero and that at pH = 7 T1/2 is not zero. So I trust these data more and will use them later on.

The fact that the stability of cyanohydrins goes asymptotically to zero above pH = 7 seems for laymen a contradiction with the reaction times of Borax catalysis at pH = 9 – 11 (1 – 8 hours). But its not, because boron will form a complex with mandelonitrile, raising the T1/2 for the hydrolysis leading to HCN + benzaldehyde.

The formation constant Kform of mandelonitrile = 210 and for butanone cyanohydrin = 38.17 I think that the Kform of acetone cyanohydrin comes close to the last one, because these are of course very similar compounds. The stabilities in water for mandelonitrile are then more like the ones found for acetone cyanohydrin multiplied by 210 / 38. From the stability data of acetone cyanohydrin in water there follows a T1/2 = 3.0 – 1.6 minutes for pH = 7.1 – 7.4, the physiological pH range of mammals. Because the data is at room temperature I presume (293 K) this must be calculated to 310 K. This gives 0.92 – 0.49 minutes for the T1/2 range.5 Then the data must be corrected for the difference in formation constant this gives 5.0 – 2.7 minutes for the T1/2 of mandelonitrile chemical hydrolysis at physiological pH.

 

1.3. Metabolism of amygdalin when mammals consume bitter almonds

This is a complex issue because there are amygdalin attacking enzymes present in the bitter almonds, in the intestines, in the liver and in all other mammal cells. The presence of these enzymes in bitter almonds is part of the already mentioned defence mechanism. When the bitter almonds are disrupted by chewing, the amygdalin and enzymes come together and already in the mouth there is metabolism in the saliva. This first step of the total metabolism is given in Fig. 6.

 

1. = gentiobiose-O-CH(Ph)-CN = glucoside-glucoside-O-CH(Ph)-CN

1. ß beta-glucosidase à glucose + 2. ß beta-glucosidase à glucose + 3.

(Or 1. ß beta-glucosidase à glucoside-glucose + 3.)

ß cyanohydrin lyase à HCN + Ph-CH=O

 

Fig. 6: Proposed enzymatic hydrolysis of amygdalin in the mouth after chewing the bitter almonds. The enzyme complex of beta-glucosidases and cyanohydrin lyase is called emulsin complex. Note the pH of (human) saliva is about 7.1.18

 

According to the pH of saliva, which is the medium in the mouth, there may be no need for the cyanohydrin lyase. The hydrolysis of mandelonitrile, releasing HCN, maybe goes even faster by the chemical route. But I think that the lyase is crucial to the defence mechanism when the pH < 7, which can occur during microbial attack of the bitter almonds by Fungi (Optimum growth pH around 5).19 It can also be that the contents of the chewed bitter almonds exceeds the buffer capacity of the saliva in the mouth. And that the pH < 7.1. As a matter effect the optimum pH and temperature for the enzymes in the emulsin complex is 5.8 and 50oC (beta-glucosidase from maize) 5.7 and 25oC (mandelonitrile lyase in almond).20 I do not think it is a coincidence that the optimum pH’s of both enzymes are so close together! Also the fact that they are below 7 is no coincidence. Remember the defence mechanism is developed by evolution. And adapted to the most occurring circumstances of attack.

But how fast is the biochemical process given in Fig. 5? Fortunately measurements are done on a very similar model. Hanssen et al. describe that amygdalin was totally hydrolysed to glucose and mandelonitrile in three hours at room temperature with the emulsin complex in grounded sweet almonds. The pH range they tested was 5.5 – 6.5.21 This can only mean that not all the HCN bonded in amygdalin is released by chewing bitter almonds in the mouth. One chews maximal for a minute before swallowing. The pH range of Hanssen overlaps the optima of the enzymes from the emulsin complex.

 

1. ß H+ à Sugar-OH+-CH(Ph)-CN ß à Sugar+ + HO-CH(Ph)-CN ß H2O à Sugar-OH2+

ß à Sugar-OH + H+

 

Fig. 7: The acid hydrolysis of the glucosidic bond to the mandelonitrile moiety in amygdalin in chyme.

 

In Fig. 6 is the acid catalysed hydrolysis shown of amygdalin. Of course this reaction can also happen with prunasin. As a matter of fact by hydrolysing a glucosidic bond amygdalin can chemically be converted to prunasin and from there to mandelonitrile and glucose. This reaction will certainly happen in the stomach. There is no way that by this mechanism some how HCN is released.

Amygdalin and prunasin are chemical insensitive to mild alkaline conditions which can occur in the mammal body.17

Most likely the beta-glucosidases and cyanohydrin lyase from the emulsin complex do not pass the strong acidic stomach barrier. Here is also the action of the protein splitting enzyme pepsin. But there are also beta-glucosidase producing micro-organisms in the intestines to consider.

 

Part of the intestine

Medical terms

Approximate pH

First part small intestine (duodenum)

Pylorus to jejunum

6.0

Second part small intestine (jejunum)

Duodenum to ileum

7.5

Colon

Ileum to rectum

6.8

 

Table 1: pH’s of the several parts of the intestine tract.

 

It is difficult to say what happens when the almond pulp containing chyme comes in to the duodenum. Certainly it is "neutralised" by pancreatic bile to a pH of about 6.0. (See Table 1) This is near the pH optimum of the beta-glucosidases and cyanohydrin lyase of the emulsin complex. But I am not sure if any enzyme molecules will have survived the action of pepsin in the stomach. Which makes it even more complicated is that there is also an abundance of beta-glucosidases from micro organisms present in the duodenum and from production by the small intestine cells it self (cytosolic beta-glucosidase22). I do not know the optimum pH’s so I can only speculate that these beta-glucosidases are different in amino acid sequence and thus specifications from those of the emulsin complex, but that they are very active in the duodenum because else what are they doing there! All one can say is that the possible products of the chemical and biochemical hydrolysis in mouth, stomach chyme and duodecum of amygdalin at pH = 6.0 up to this point are CN-, HCN, benzaldehyde, mandelonitrile, mandelic amide, mandelic acid, glucose, gentiobiose, prunasin and unreacted amygdalin.

All these compounds can be absorbed through the duodecum wall in to the blood stream, which has a pH of around 7.4 (there is only a very small tolerance of about 0.1 units).

Can amygdalin and prunasin be substrates for an SN1 or SN2 kind of reaction in the blood stream releasing cyanide ions or mandelonitrile? I do not think so because:

·         The pH is only 7.4. So the hydroxide ion concentration is low.

·          CN- is a stronger nucleophile than OH-. Nucleophilicities of CN- is 5.1 and for OH- the value is 4.2.27 Glucoside-O-CH(Ph)-CN + OH- à Glucoside-O-CH(Ph)-OH + CN-

·         Glucoside-O- is a very poor leaving group.11 Glucoside-O-CH(Ph)-CN + OH- à Glucoside-O- + HO-CH(Ph)-CN

In other words the releasing of mandelonitrile from amygdalin and prunasin can only be done biochemical and is not spontaneous (means chemical).

Since almost 100% of the substances absorbed by the bloodstream is done in the small intestine, the chemistry in the colon is not relevant here. But I can say that the processes in the duodecum are the same for the jujuneum and on a very small scale for the colon.

 

1.4. Detoxification of HCN in mammals

The detoxification process of HCN in the mammal body is excellently described in the following reference7.

How does the cyanide do its deadly work in cells? Well you can read that on the following references7,16

 

2. Materials and Methods

The computer models were originally programmed in PASCAL and the results were checked by writing a similar program in VISUAL BASIC. The input of the programs comes either directly from literature data or was calculated from literature data. The PASCAL program was a console application with can be run from the DOS prompt of Windows XP. The computer was a Pentium-4 2.8 GHz 512 Mb RAM. The VISUAL BASIC program was a GUI with needed .NET Framework to run on an AMD Athlon 700 MHz 128 Mb RAM with Windows 2000. Graphs from Fig. 8 and 9 where plotted by means of a VISUAL BASIC application. All VISUAL BASIC applications where programmed with VISUAL STUDIO 2003 .NET.

 

3. Results

3.1. The LD50 (oral, rat) bitter almonds. First Estimation

This First Estimation is based on the fact that about 8.0 % of the HCN is released by chewing the bitter almonds in the mouth4,5 Despite laymen opinions no HCN can be released from amygdalin, prunasin and free mandelonitrile in the strong acidic environment of the stomach. It is highly likely that there is also no beta-glucosidase and cyanohydrin lyase activity here. It might even be that the protein splitting enzyme pepsin destroys all these enzymes of the emulsin complex found in bitter almonds.

 

mmol / Kg HCN released in the mouth = 0.080 * x (x = number of mmol amygdalin / Kg in the feed)

mmol / Kg HCN lethal LD50 (oral, rat) = 0.81 / 27 = 0.030

mmol / Kg amygdalin lethal = 880 / 457 = 1.923

mmol / Kg amygdalin processed by the rest of the body = 0.92 * x

8.0 % of the HCN in amygdalin is released in the mouth. This is 0.080 * x (A)

0.030 * x / 1.93 = (1.6 % (of x) = molar ratio LD50 HCN / LD50 amygdalin) The remaining part amygdalin which is transformed by the body contributing to a deadly dose HCN. (B) This is an approximation because it does not take in to account that HCN is transformed to thiocyanate by the body.

The total A + B must be the same as the LD50 (oral, rat) HCN = 0.030 mmol / Kg

0.080 * x + 0.030 * x / 1.9 = 0.030 => 2.7 * x + 0.52 *x = 1 => x = 1 / 3.2 = 0.31 mmol / Kg

This brings us to a 0.31 * 457 = 140 mg / Kg the LD50 (oral, rat) for amygdalin in bitter almonds, with the presence of beta-glucosidases and cyanohydrin lyases.

And this brings us to a 140 / (0.051 * 1000) = 2.7 g / Kg LD50 (oral, rat) for bitter almonds.

Because 1.6% of the amygdalin is transformed to HCN this gives an approximate Time Of Death of 19 + 1 = 20 minutes4,5

 

Scheme 1: The First Estimation of the LD50 (oral, rat) bitter almonds.

 

The LD50 (oral, rat) 880 mg pure amygdalin / Kg body weight in Scheme 1 comes from Adewusi et al.23

 

3.2. The LD50 (oral, rat) bitter almonds. The Computer Models 1 and 2

There is great difference in the HCN LD50-values from several sources on the internet. Table 1 gives the values of some of the sources.

 

Mammal

Reference [7]

Reference [16]

Reference [24]

Mouse

0.99

 

 

Rat

0.81

4.21

8.5

Guinea pig

1.43

 

 

Rabbit

0.66

2.48

 

Cat

0.81

 

 

Dog

1.34

 

 

Monkey

1.30

 

 

 

Table 2: The LD50 (oral, mammal) HCN values in mg / Kg body weight from several sources from the internet.

 

Since HCN has a pKa of 9.22 virtually all cyanide is in the protonated form (HCN) active in the mammal body.16 So we do not need to worry about the toxicity of the cyanide ion.

It happens to be so that I have the most kinetic data about rats. So my Computer Models will describe the some of the biochemical and chemical processes in the rat. About Table 2: I do not trust the data from row16 very much and row24 even less. To be on the save side I choose the minimum value of 0.81 mg / Kg body weight of rat and put that in the models. This is also data that comes with data from other mammals, and they are more or less of the same magnitude.

I now will describe the used input in model 1 and 2:

·         The LD50 (oral, rat) amygdalin, WITHOUT the emulsin complex, comes from Adewusi et al.23. This figure is used to calculate the T1/2 amygdalin in the body by model 1.

·         The T1/2 HCN in the body of 14.1 minutes is the rate at which cyanide is eliminated in the blood of rats after a single oral dose. This then takes the reaction of HCN with biomolecules, the transformation to thiocyanate by rhodanese in liver and other body parts, the exhalation of HCN by the lungs and excretion of HCN in the urine and faeces in to account.

·         The T1/2 excretion amygdalin in urine was calculated by finding the average of 11.7% (200 mg intake), 18.5% (400 mg intake) and 12.4% (600 mg intake) from Adewusi.23 This gives 14.2% average in 48 hours. This leads to a T1/2 of 13,000 minutes. This is of course only a minor process and will not affect the significant digits of the results. (Which are actually given in two or three significant digits depending whether I wanted subtlety.) No nitriloside was ever detected in the faeces of rats.7

·         The T1/2 amygdalin in the mouth, emulsin complex present, (8.3 minutes) is calculated from literature data of Hansen et al..4,5,21 Since this is a very low value and model 2 assumes that the reaction leads directly to HCN, this value must have the greatest impact on the results although the chewing time is fixed to only 1 minute. This is perhaps the greatest source of error in the results because Hansen states that is data specifically for the reaction to mandelonitrile. And the chemical reaction of mandelonitrile to HCN has an T1/2 of about 5 minutes as I calculated in paragraph 1.2. at pH = 7.1 of the saliva. What also must be taken in to account is the fact that besides the chemical hydrolyses of mandelonitrile there is also the action of the emulsin complex cyanohydrin lyase. So the T1/2 mandelonitrile is smaller than 5 minutes. Nevertheless the results obtained by assuming all these reactions combined to have an T1/2 = 8.3 minutes will lead to an to low value of the results. Considering this: It is also assumed that the concentrations of enzymes in the mouth is the same as in the experiments of Hansen. This may not be the case. Another source of uncertainty in the results.

 

 

Input

Value

Result

Value of estimation

First Estimation

LD50 (oral) amygdalin

880 mg / Kg body weight

LD50 (oral) bitter almonds

2.7 g / Kg body weight

 

LD50 (oral) HCN

0.81 mg / Kg body weight

LD50 (oral) amygdalin with emulsin complex present

140 mg / Kg body weight

 

T1/2 amygdalin in the mouth

8.3 minutes

Time of death

20 minutes

 

T1/2 amygdalin in the body

806 minutes

 

 

 

 

 

 

 

Model 1

LD50 (oral) amygdalin with NO emulsin complex present

880 mg / Kg body weight

LD50 (oral) bitter almonds

17 g / Kg body weight

 

T1/2 HCN in the body

14.1 minutes

T1/2 amygdalin in the body

806 minutes

 

LD50 (oral) HCN

0.81 mg / Kg body weight

Time of death

77 minutes

 

T1/2 excretion amygdalin

13,000 minutes

 

 

 

 

 

 

 

Model 2

T1/2 amygdalin in the mouth

8.3 minutes

LD50 (oral) bitter almonds

3.3 g / Kg body weight

 

T1/2 amygdalin in the body

806 minutes

LD50 (oral) amygdalin with emulsin complex present

170 mg / Kg body weight

 

T1/2 HCN in the body

14.1 minutes

Time of death

61 seconds

 

LD50 (oral) HCN

0.81 mg / Kg body weight

 

 

 

T1/2 excretion amygdalin

13,000 minutes

 

 

 

Table 3: The results for the First Estimation and the two Computer Models. Simulations of the chemical and biochemical processes in rats.

 

 

Fig. 8: Graph made from the logfile of Computer Model. Intake amygdalin is 880 mg / Kg body weight (1.9 mmol / Kg bodyweight) Peak [HCN]body at 77 minutes and 0.81 / 27 = 0.03 mmol / Kg body weight. NO BETA-CLUCOSIDASES IN THE FEED!

 

 

Fig. 9: Graph from the logfile of Computer Model 2. Intake amygdalin is 170 mg / Kg body weight (0.37 mmol / Kg). Peak [HCN] at 1.01 minutes and 0.81 / 27 = 0.03 mmol / Kg body weight. WITH BETA-GLUCOSIDASES IN THE FEED like with bitter almonds.

 

4. Discussion

The objective of this article comes from the fact that Adewusi published the LD50 (oral, rat) amygdalin = 880 mg / Kg body weight and a LD100 (oral, rat) amygdalin with beta-glucosidases present = 600 mg / Kg body weight.23 But he forgot to measure the LD50 amygdalin with beta-glucosidases present like in bitter almonds! Since I cannot do the experiment I decided to estimate the value by means of a computer model.

The calculated LD50 (oral, rat) of amygdalin (First Estimation) in bitter almonds of 140 mg / Kg is indeed smaller then the value of a LD100 (oral, rat) of 600 when beta glucosidases are present of Adewusi. This makes my estimation reliable.

The computer model 2 also finds a value that is lower than the LD100 value of Adewusi. LD50 (oral, rat) amygdalin, emulsin complex present, is found to be 170 mg / Kg. The fact that this number is very close to the First Estimation value of 140 makes these independent found data reliable.

The estimated T1/2 amygdalin in the body of 806 minutes takes all the reactions that lead to HCN in to account. Including:

·         Biochemical amygdalin to prunasin + glucose (beta-glucosidase).

·         Biochemical amygdalin to mandelonitrile + gentiobiose (beta-glucosidase).

·         Biochemical prunasin to mandelonitrile + glucose (beta-glucosidase)

·         Biochemical mandelonitrile to benzaldehyde + HCN (cyanohydrin lyase)

·         Chemical mandelonitrile to benzaldehyde + HCN (physiological pH = 7.1 – 7.4).

Since also benzaldehyde is formed we need to check what the LD50 (oral, rat) for this compound is. Maybe rats consuming bitter almonds die from benzaldehyde poisoning and not from the HCN. Or perhaps from both! The LD50 = 1300 mg / Kg = 12 mmol / Kg.25 So HCN is 410 times more toxic to rats than benzaldehyde. So we do not need to worry about benzaldehyde because it is formed in an equimolar amount as HCN.

The differences in the Time Of Death values can be explained by the fact that First Estimation assumes 1.6% elimination of amygdalin in the body. In Model 2 demonstrates that this value is much to high it is more likely that the Time Of Death follows right after swallowing the toxic brew formed in the mouth when rats eat about 3 g bitter almonds all at once. I assumed that the chewing time is about 1 minute.

Using also T1/2 values for amygdalin to mandelonitrile biochemical and for mandelonitrile to HCN biochemical and chemical in the mouth (I presumed that these are all incorporated in the 8.3 minutes from Adewusi23) will most likely give a higher result for the LD50. Or it should be that the biochemical reaction is much faster than the chemical hydrolyses of mandelonitrile

 

5. Conclusions

·         T1/2 amygdalin in the mouth of mammals leading directly to HCN at a pH of about 6 and 310 K is appoximately 8.3 minutes according to simple calculations.4,5 This maybe the greatest source of error in the results.

·         T1/2 amygdalin in the body of rats is about 810 minutes according to Computer Model 1. 50% of rats fed 880 mg pure amygdalin / Kg body weight will die within 77 minutes. The concentration HCN in the body is then at its peak in Model 1. Solely because of the transformation of amygdalin to HCN by the biochemical and chemical processes in the body. This number assumes that there is no release of HCN in the mouth because of lack of emulsin complex. But maybe there are beta-glucosidases in the saliva, then there could be (only) chemical hydrolysis of mandelonitrile. Since I do not think that there are cyanohydrin lyases in the saliva.

·         The LD50 (oral, rat) amygdalin with emulsin complex present = 140 – 170 mg / Kg body weight according to the First Estimation and Computer Model 2.

·         The LD50 (oral, rat) bitter almonds = 2.7 – 3.3 g / Kg body weight according to First Estimation and Computer Model 2.

·         The potential availability of amygdalin to the body cells is 92% of the intake. Since only 8% is hydrolysed in the mouth taken an average chewing time of 1 minute4,5

·         I am of course curious to what the experimental results (LD50) would be of feeding rats bitter almonds. Maybe this article will inspire scientists to do this experiment.

·         According to literature data there is no fear for chronic poisoning effects of feeding mammals high nitriloside diets.16 The toxicity of the formed thiocyanate is significantly less than that of cyanide, but chronically elevated levels of blood thiocyanate can inhibit the uptake of iodine by the thyroid gland, thereby reducing the formation of thyroxine.24 Maybe giving extra (salt with) iodine in the feed can compensate this.

·         According to my models and calculations bitter almonds are just as toxic as common kitchen salt. LD50 (oral, rat) salt = 3 g / Kg body weight. But I also learned from the Toxicology newsgroup that estimations and predictive methods are no substitute for the values found by experiment.

·         More (literature) research is needed to say what a good estimation is of the LD50 (oral, rat) HCN because I only found contradictionary information in the literature from the internet. This figure, a key input of my models, is crucial for the accuracy of the results.

·         The computer models can be refined by taking the following processes in to account:

1.      T1/2 amygdalin to prunasin + glucose in the mouth (emulsin complex).

2.      T1/2 amygdalin to mandelonitrile + gentiobiose in the mouth (emulsin complex).

3.      T1/2 prunasin to mandelonitrile + glucose in the mouth (emulsin complex).

4.      T1/2 mandelonitrile to HCN + benzaldehyde in the mouth (emulsin complex and chemical).

5.      T1/2 HCN to formic acid in the stomach (chemical hydrolysis).

6.      T1/2 mandelonitrile to mandelic acid in the stomach (chemical hydrolysis).

7.      T1/2 amygdalin to prunasin + glucose in the body (beta glucosidases).

8.      T1/2 amygdalin to mandelonitrile + gentiobiose in the body (beta-glucosidases).

9.      T1/2 prunasin to mandelonitrile + glucose in the body (beta-glucosidases).

10.  T1/2 mandelonitrile to HCN + benzaldehyde in the body (cyanohydrin lyases and chemical).

11.  T1/2 excretion prunasin in the urine.

12.  T1/2 exhalation HCN by the lungs.

13.  T1/2 excretion mandelonitrile in the urine.

14.  T1/2 excretion HCN in the urine.

·         These are of course only the major processes with can be refined even more by separating processes in the small intestine, colon, blood, liver, kidneys etc. The processes mentioned are the ones I think one can find literature data about and incorporating them in to a computer model would give a more precise number of the LD50 (oral, mammal) bitter almonds. A good model should also take the concentration of the enzymes in to account and maybe use Michaelis-Menten Kinetics instead of the T1/2 approuch.

 

-o0o- Please also visit: The new Journal of Randomics site and the cumulated result of the site here

 

References & Notes:

1) Krebs Jr E.T. ‘The nitrilosides in plants and animals’

http://www.navi.net/~rsc/nitrilo1.htm

2) Oke O.L. ‘The role of hydrocyanic acid in nutrition’ World Review of Nutrition and Dietetics II 170-98 (1969)

3) Gibb M, Orr R ‘Grazing behaviour of ruminants’ Igger Inovations 54-57 (1997)

http://www.iger.bbsrc.ac.uk/Publications/Innovations/in97/Ch9.pdf

4) The T1/2 amygdalin can be calculated from the data of Hansen > 99% decay in 3 hours at room

temperature. Then the formula Nt = N0 * exp(-0.693 * t / T1/2) => 1 =

100 * exp(-0.693 * 180 / T1/2) => T1/2 = -180 * 0.693 / ln(1 / 100) = 27 minutes.

Exp(x) = ex and Ln(2) = 0.693… So: Nt = N0 * 2^(-t / T1/2)

Sheep transform 95% of the amygdalin to HCN in one hour. T1/2 = 14 minutes.

5) The rate of a chemical reaction increases with temperature as such that with every 10 K

increase the rate doubles. So the T1/2 can then be halved. T1/2 amygdalin with emulsin complex

present (like in bitter almond pulp-water mixtures) = 27 minutes at 293 K (room temperature)

à 27 * 2-17/10 = 8.3 minutes at 310 K (body temperature). Anonymous ‘ The effect of temperature on reactions rates’

http://www.chemguide.co.uk/physical/basicrates/temperature.html

6) Anonymous ‘Sheep’ http://www.ics.uci.edu/~pazzani/4H/Sheep.html

7) Anonymous ‘Opinions of the scientific panel on food additives, flavorings, processing aids and materials in contact with food (AFC)’ The EFSA Journal 105 1-28 (2004)

http://www.efsa.eu.int/science/afc/afc_opinions/698/afc_opinion21_ej105_hydroacidinflav_en1.pdf

8) Russell K., Kaur R. ‘Almonds’

http://www.foodscience.ac.nz/research_topics/nuts/almonds.htm

9) Anonymous ‘Almonds’ Transport Information Service

http://www.tis-gdv.de/tis_e/ware/nuesse/mandeln/mandeln.htm

10) Anonymous ‘Opinion of the scientific committee on animal nutrition on undesirable substances in animal feed’

http://europa.eu.int/comm/food/fs/sc/scan/out126_bis_en.pdf

11) March J. ‘Advanced organic chemistry’ Wiley Intersience (1984)

12) Ziegler T., Hörsch B., Effenberger F. Synthesis, 575-578 (1990).

13) Boudinot S ‘Anatomy of the gastrointestinal tract and phenobarbital absorbtion’

http://chemcases.com/pheno/pheno14a.htm

14) Anonymous ‘Chapter 2 Secondary phosphine oxide (SPO’s) and nitrile hydrolysis’

http://dissertations.ub.rug.nl/FILES/faculties/science/2004/x.jiang/c2.pdf

15) Anonymous ‘Acetone cyanohydrin’

http://www.chem.unep.ch/irptc/sids/OECDSIDS/75865.htm

16) Simeonova F.P., Fishbein L., ‘Hydrogen cyanide and cyanides human health aspects’ IPCS INCHEM

http://www.inchem.org/documents/cicads/cicads/cicad61.htm#6.0

17) Ternay A. L. ‘Contemporary organic chemistry’ WB Saunders Company (1979)

18) Tunsky G. ‘What the cell is going on? The battle for health is over pH’

http://www.jaynesgarden.com/Body/body-ph.htm

19) Anonymous

http://www.bae.uky.edu/~snokes/BAE549thermo/microbio/growthfactors.htm

20a) Anonymous ‘Uniprotkb/Swiss-prot entry P49235’

http://www.expasy.org/uniprot/P49235

20b) Anonymous ‘R-oxynitrilase’

http://www.x-zyme.com/s-oxynitrilase.htm

21) Hanssen E., Sturm W. ‘Uber Cyanwasserstof in Prunoideensamen und einigen anderen Lebensmitteln. 1. Mitteilung Bestimmung glucosidisch gebundener Blausäure an bitteren Mandeln.’ European Food Research and Technology 134(2) 69-80 (1967)

22) Anonymous ‘Human protein reference database’

http://www.hprd.org/protein/09428?selectedtab=SUMMARY

23) Adewusi S.R., Oke O.L.’On the metabolism of amygdalin. 1. The LD50 and biochemical changes in rats’ Can. J. Physiol. Pharmacol. 63(9) 1080-1083 (1985)

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&list_uids=2932206&dopt=Abstract

24) Faust R.A. ‘Toxity summary for cyanide’ Oak Ridge National Laboratry (1994)

http://risk.lsd.ornl.gov/tox/profiles/cyanide.doc

25) Anonymous ‘Benzaldehyde’ ChemicalLAND21

http://www.chemicalland21.com/arokorhi/specialtychem/perchem/BENZALDEHYDE.htm