Monoterpenes are important constituents of many essential oils and are used extensively as flavouring agents in the food industry. Citral and linalool, both acyclic monoterpenes have been shown to cause peroxisome proliferation in the liver of rats (Jackson et al.,1987; Roffey et al., 1990).

Peroxisome proliferators like clofibrate and many other hypolipidaemic agents when given to rodents result in an increase in the transcription of genes for proteins involved in peroxisome ß-oxidation of long-chain fatty acids and in the morphologically detectable proliferation of peroxisomes Reddy et al., 1983; Hawkins et al., 1987). This phenomenon was also observed in rodents fed a diet rich in long-chain fatty acids but with a smaller response (Neat et al., 1980).

These agents have been shown simultaneously to enhance transcription of the gene for cytochrome P450IVA1 and increase the cellular content and activity of this enzyme (Gibson et al., 1982; Gibson, 1989).

Based on these findings and studies conducted on a diverse number of hypolipidaemic agents, a correlation was proposed between the ability of various agents to induce cytochrome P450IVA1 and cause peroxisome proliferation (Sharma et al., 1988; Lake et al., 1984). A number of mechanisms were proposed to explain this correlation.

Sharma et al. (1988) proposed that the induction of cytochrome P450IVA1 by hypolipidaemic agents increases fatty acid w-hydroxylation and the formation of fatty dicarboxylic acids which cannot readily be metabolised by the mitochondria, therefore, the peroxisome is presented with a substrate overload which adapts by proliferation in order to chain shorten these dicarboxylic acids for subsequent mitochondrial metabolism.

Reddy et al. have proposed that peroxisome proliferators exert their effects by a ligand-receptor mediated mechanism (Reddy and Lalwani, 1983; Rao and Reddy , 1987; Reddy and Rao, 1986; Lalwani et al., 1987) which causes both the induction of peroxisome ß-oxidation and transcription of the gene coding for the enzyme. Peroxisome proliferators are structurally diverse, therefore, a receptor theory could only be possible if:

Recently, a mouse receptor (mPPAR, for mouse peroxisome proliferation activated receptor) has been isolated that is activated by peroxisome proliferators (Issemann and Green, 1990; Dreyer et al., 1992).

Lock et al. (1989), link inhibition of the mitochondria into the correlation.

Administration of DEHP results in a transient accumulation of small droplets of neutral lipid in the liver (Mann et al. , 1985; Mitchell et al., 1985) which disappear when the peroxisomes are induced. The long-chain dicarboxylic acids (LCDCA) are also able to stimulate 3H-thymidine incorporation into DNA in isolated rat hepatocytes in culture (Lock et al., 1989).

In some cases, a correlation does not exist (Roffey et al., 1990). Roffey et al. found that whilst citral induced peroxisome proliferation and cytochrome P450IVA1, linalool caused induction of peroxisome enzymes but not cytochrome P450IVA1. Lazarow et al. (1982), observed that while bezafibrate does induce peroxisome ß-oxidation enzymes it does not elicit peroxisome proliferation in rat.

Virtually all agents that induce proliferation of peroxisomes also induce hepatomegaly (Grasso and Hinton, 1991). Studies with fenofibrate (Price et al., 1986), trichloroethylene (Elcombe et al., 1985) and di-(2-ethylhexyl)phthalate (Mitchell et al. , 1985) have shown that peroxisome proliferators and liver growth correlate and are dose-dependent. The liver enlargement and increased enzyme activity regress on cessation of treatment (Elcombe et al., 1985). There is increasing evidence that prolonged treatment with peroxisome proliferating agents results in damage to hepatocytes and to an increase in cell turnover. Cells in mitosis are particularly susceptible to damage. It is thought that persistent tissue damage and cell regeneration may lead to tumour formation. Peroxisome proliferators, in addition also stimulate cytosolic hydrolase (Meijer and DePierre, 1987) in mice which may form diols from fatty acid epoxides due to the action of H2O2. They inhibit glutathione peroxidase and superoxide dismutase activity (Ciriolo , 1982), thereby, preventing the efficient scavenging of hydroperoxides. The persistent alteration in the levels of these enzymes may contribute to the development of tumours in the liver.

The present study is designed to find the effects of 10 cyclic monoterpenes on the cytochrome P450 family (P450IA1, P450IA2, P450IIE, P450III, P450IVA1) and cyanide-insensitive palmitoyl-CoA oxidation as an indication of peroxisome ß-oxidation enzyme activity. The compounds found to increase these activities can then be studied at a later date on high dose chronic administration to see if they will be responsible for carcinogenesis, and because most peroxisome proliferators are threshold dependent, can be extrapolated to man, in order to find an ADI level.


Peroxisomes are single membrane bound cytoplasmic organelles present in all mammalian cells except red blood cells. However, their number, size, and enzyme profile varies between different tissues (de Duve, 1983). Biochemical analysis revealed a content of several oxidative enzymes. The feature common to all these enzymes was that they generated hydrogen peroxide (H2O2) as a product of their reaction. Thus, it was considered that the peroxisome had developed in the course of evolution in order to protect the cell from the potential harmful effects of these particular oxidative enzymes. The reactions are catalysed by the appropriate FAD - linked enzyme:

  1. urate oxidase,
  2. xanthine oxidase and
  3. L- and D- amino acid oxidases.
The potent oxidizing agent, hydrogen peroxide, which all three reactions generate, is rendered harmless by the highly active haem-containing catalase (enzyme 4). In rat liver the peroxisomes have a characteristic paracrystalline core, which consists of the enzyme urate oxidase.

However, the peroxisomes of other animal organs and plant tissues lack this core, and so they are difficult to distinguish from the many other membrane-bound vesicles seen in electron micrographs unless histochemical staining (for example catalase) is used. One feature of peroxisomes is that they are inducible and, therefore, an administration of the hypolipidaemic drug clofibrate to rats leads to a tenfold increase in number of peroxisomes per liver cell. In certain yeasts grown on methanol, the peroxisomes can occupy almost the whole volume of the cytoplasm. Peroxisomes are involved in lipid degradation and biosynthesis. A ß-oxidation pathway similar in function to that of mitochondria is present in peroxisomes. However, the peroxisomal pathway differs from that of mitochondria in several important respects:

  1. mitochondrial acyl-CoA oxidation is dependent upon the presence of carnitine, since the formation of acyl carnitine catalysed by carnitine acyl transferase is an absolute requirement for the transport of the fatty acid moiety through the inner mitochondrial membrane. A carnitine dependent transport system for acyl-CoAs does not exist in peroxisomes,
  2. it can catalyse the degradation of very-long-chain fatty acids (C20 or greater),
  3. it does not act on short fatty acids (C6 or less),
  4. the first enzyme in the pathway is an FAD-linked acyl CoA oxidase, rather than the FAD-linked dehydrogenase of the mitochondrial pathway,
  5. the first reaction is coupled to the formation of H2O2,
  6. the other enzymes are distinct gene products specific to the peroxisome.
In the peroxisome ß-oxidation pathway, a fatty acid (usually C8 - C22 or more) is activated to an acyl-CoA derivative by an ATP-dependent acyl-CoA synthetase located in the peroxisome membrane. All subsequent reactions take place in the peroxisome matrix. The fatty acyl-CoA is reduced with the utilization of oxygen to trans-2-enoyl-CoA, yielding hydrogen peroxide; this reaction is catalysed by acyl-CoA oxidase, the rate limiting enzyme of peroxisomal ß-oxidation. The next two reactions are catalysed by a bifunctional protein possessing the activities of enoyl-CoA hydratase and 3-hydroxy acyl-CoA dehydrogenase. The final reaction of ß-oxidation is carried out by 3-ketoacyl-CoA thiolase which cleaves 3-ketoacyl-CoA into acetyl-CoA and a saturated acyl-CoA with two carbons less than the original molecule. The newly formed acyl-CoA then re-enters the ß-oxidation pathway. Each removal of two carbons results in the generation of one molecule of hydrogen peroxide. The components of the peroxisomal ß-oxidation system differ significantly from enzymes of mitochondrial fatty acid ß-oxidation with respect to their molecular and catalytic properties. For example, peroxisomal ß-oxidation does not require carnitine and is not inhibited by potassium cyanide.

Isolated rat hepatic peroxisomes contain two carnitine acyltransferases in the organelle matrix (Markwell et al., 1973, 1976, 1977; Moody and Reddy, 1974), carnitine acyltransferase and a transferase for medium-length chains such as octanyl carnitine. Their substrate specificity suggests that the C2 and C8 acyl-CoA products of hepatic peroxisomal ß-oxidation are converted to the corresponding carnitine derivatives and then passively diffuse out of the peroxisomes into the cytoplasm. Thus peroxisomal ß-oxidation may provide acetyl units for other cellular synthesis, or the acyl carnitine products may be further oxidised in the mitochondria.

The question to be addressed is, why is there the need for the carnitine acyltransferase, if the fatty-acyl-CoA can passively diffuse into and out of the membrane? Mention has not been made of acylcarnitine translocase which functions as the carrier for acylcarnitine. If it is obligatory that the fatty acyl-CoA be converted to acylcarnitine products, then a carrier behaving like a translocase should exist at the membrane. Inhibiting either the acylcarnitine transferase or the translocase should inhibit fatty acid uptake by mitochondria, overload the peroxisome with substrates for its ß-oxidation pathway, inhibit either the acylcarnitine transferase at the peroxisome or the translocase which then cannot transport products out, therefore, there will be swelling of the peroxisome since the fatty acyl-CoA can easily diffuse into the matrix, and as a homeostatic mechanism result in peroxisome proliferation in order to compensate for the overload, leading to a futile cycle, since the products cannot be transported out, unless a mechanism comes into play to make the membrane more permeable, leading to diffusion of not only the products (fatty acids) but also H2O2.

Cytochrome P450IVA1

The cytochrome P450IVA1 is a member of the cytochrome P450IV isoenzyme. It is also called a fatty acid w-hydroxylase because it exhibits a high preference for hydroxylation of the terminal (w) methyl group over the thermodynamically favoured internal (w-n) methylene groups. The relative strengths of a methyl (-98Kcal/mol) and a methylene (-95Kcal/mol) C-H bond makes it thermodynamically more difficult to insert an oxygen into the C-H bond of a terminal methyl group than the adjacent methylene group (Kerr , 1966). The oxidation of fatty acids by cytochrome P450 enzymes not specifically designed to carry out w-hydroxylations yields primarily w-1 and w-2 hydroxylated products (Tanaka et al., 1990).

The fatty acid w-hydroxylase represents approximately 1-2% of the cytochrome P450 in uninduced rat liver microsomes and 16-30% after induction with clofibrate (CaJacob et al., 1988). The preference for w-hydroxylation may be due to a limited substrate access to the ferryl oxygen at the active site.

Some differences have been found between the members of the P450IV family in relation to substrate specificity and their tissue expression. The liver P450IVA1 is active in the w-hydroxylation of lauric acid but does not metabolise prostaglandins, whereas the lung P450IVA1 catalyses the w-hydroxylation of several prostaglandins but not lauric acid (Tamburini et al., 1984; Matsubara et al., 1987). The liver P450IVA1 exhibits a narrow substrate specificity by metabolizing both lauric acid and arachidonic acid relatively well (Bains et al., 1985) but not the prostaglandins (Matsubara et al., 1987).

Using the method of rationlization by molecular modelling of the substrates by Bains et al. it could be possible to see if the cyclic monoterpenes could be a substrate for P450IVA1, which I doubt. Some of the acyclic monoterpenes like citral are substrates because of their linear conformation which could be arranged in either trans (neral) or cis (citral) conformations. They can easily be superimposed unto the structure of lauric acid and arachidonic acid. The cyclic monoterpenes might possibly be substrates for the P450IVA4 since this enzyme prefers the prostaglandins which have a central cyclopentane ring. If metabolism or some favourable reaction conditions does open the ring of the cyclic monoterpenes, then it could be possible for them to become substrates for the P450IVA1 (wishful thinking). Twenty eight cyclic terpenes were studied using COMPACT computer graphic studies and none was found to be likely substrates for either P4501 or P450IIE (Lewis et al. , 1992).

Cyclic Monoterpenes

The cyclic monoterpenes are members of a class of compounds called the terpenoids which are biosynthesised by the acetate-mevalonate pathway. Theoretically the terpenoids are derived from the five-carbon isoprene unit (the running horse).

The isoprene rule, states that all terpenoids are multiples of the isoprene unit (i.e. C10, C15, C20, etc.), although this is not strictly obeyed by natural products. Acetyl-CoA is the source of all the carbon atoms in Terpenes. ß-hydroxy-ß-methylglutaryl-CoA is converted to mevalonate in a 2-stage reduction by NADPH catalysed by ß-hydroxy-ß-methylglutaryl-CoA reductase. In the second stage, mevalonate is phosphorylated by ATP to form several active phosphorylated intermediates. By means of a decarboxylation, the active isoprenoid unit, isopentenylpyrophosphate is formed. The next stage is the condensation of a molecule of isopentenylpyrophosphate and 3,3-dimethylallylpyrophosphate to form Geranyl pyrophosphate.


Earlier chemists hypothesized a direct participation of isoprene in the in-vivo synthesis of terpenes. This hypothesis was supported by the possible formation of dipentene from two isoprene units by a simple Diels-Alder process and by the wide occurrence, amongst essential oils, of compounds with the dipentene structure. Dipentene found this way is racemic. It is particularly abundant in turpentine oil. Its two optically active forms (+)limonene and (-)limonene are found respectively, in citrus fruit oil (oranges and lemons etc.) and peppermint oil.

Other dipentene derivatives are widely distributed in nature. An objection to this early postulate, however, was that ISOPRENE itself did not appear to be present in nature and could only be obtained by the pyrolysis of certain monoterpenes.

The monoterpenes can occur in:

forms as hydrocarbons and as oxygenated derivatives such as; Possible pathways for the formation of the cyclic monoterpenes in the project.

An interesting stereochemical feature of many terpenes is the fact that both enantiomers (optically active isomers) exist in nature. In some cases, a plant species produces only one of the enantiomers, whereas, a different species may produce both. As with many other natural product compounds that exist in enantiomeric forms, such as alkaloids and amino acids, the physiologic responses elicited by each isomer can differ. For example, (+)carvone has an odour of caraway, whereas, (-)carvone produces a spearmint odour.

These observations lend support to the stereochemical theory of olfaction which proposes that different kinds of olfactory receptor sites are in the nose. Odorant molecules could lodge on these sites and would have shapes and sizes (varying in stereochemistry) that were complementary to the shape and size of the particular receptor. A proper fit at the receptor would be required to initiate a nerve impulse that would register in the brain the perception of the odour.



Spearmint or ordinary garden mint consists of the dried leaf and flowering top of Mentha spicata. It has stalkless leaves and flower-clusters in a slender spike. Spearmint oil is prepared steam distillation of Mentha spicata or of Mentha cardiaca and should not contain less than 55% of carvone. Oil of spearmint contains from 45 to 60% (-)carvone, 6 to 20% of alcohols, and 4 to 20% of esters and terpenes, mainly (-)limonene and cineole. The optically isomeric form of carvone, (+)carvone is found in oil of caraway and oil of dill. Carvone when present in a plant appears to co-occur with limonene.


Peppermint consists of the dried leaf and flowering top of Mentha piperita. It has almost hairless leaves and oblong heads of flower-clusters. Peppermint oil is the volatile oil distilled with steam from the fresh overground parts of the flowering plant of Mentha piperita, rectified by distillation. Contains not less than 44% menthol. American peppermint oil contains from 50 to 78% of free (-)menthol and from 5 to 20% combined in various esters such as the acetate. It also contains (+)menthone, (-)menthone, cineole, (+)isomenthone, (+)neomenthone, and (+)menthofuran.

The biogenetic arrangement for the various species of Mentha is as follows:

Peppermint oil is a common flavouring ingredient used in chewing gums (Greenhalgh, 1979), toothpaste, confectionery and pharmaceutical products.

A short-term oral toxicity study on male and female rats for 28 days demonstrated that peppermint oil causes brain lesions at doses of 40 and 100mg/kg bw/day which was confined to the white matter especially of the cerebellum (Thorup et al., 1983a). Since peppermint oil consists of all the compounds mentioned above, further studies were carried out to identify the culprit, and also find out if the oil or any of its components is mutagenic because of its extensive usage. Pulegone and menthol were inverstigated in rats (Thorup et al., 1983b), by gavage for 28 days.

Pulegone at 80 and 160mg/kg bw/day caused histopathological changes in liver and in the white matter of the cerebellum. No signs of encephalopathy was observed in rats given menthol. The toxicity of menthol was studied on four different in vitro systems (Bernson and Pettersson, 1983) covering organ, cellular and subcellular levels. 0.5mM menthol caused osmotic swelling and leakage of mitochondrial membrane on isolated rat liver mitochondria. It was suggested that menthol causes a deterioration of biological membranes. The acute oral LD50 value was reported as 3300mg/kg ( Opdyke, 1976).

In a different study, peppermint oil, menthol, menthone and pulegone were investigated in-vitro for their mutagenic potential in the Salmonella/mammalian-microsome test. Peppermint oil, menthol and pulegone were tested on salmonella tester strains TA 1537, TA98, TA1535 and TA100 at concentrations of 800, 160, 32 and 6.4µg per plate and did not demonstrate any mutagenic properties. Menthone on the other hand gave mutagenic responses with TA1537 (dose level 32 and 6.4µg per plate without S9) and with TA97 (dose levels 800, 160, 32 and 6.4µg per plate without S9 and dose levels 160 and 32µg per plate with S9). Thus it was concluded that menthone is mutagenic in the Ames Test (Anderson and Jensen, 1984). An in-vivo 28 days test on male and female rats was done with menthone at dose levels of 200, 400 and 800mg/kg bw/day. Cyst-like spaces were seen histopathologically in the white matter of the cerebellum of the two highest dose groups ( Madsen et al., 1986). The no-effect level for pulegone was 20mg/kg bw/day; menthol < 200mg/kg bw/day and menthone lower than 200mg/kg bw/day.

From an earlier study by Parke and Rahman , (1969) that terpenoids caused an induction of the drug metabolising enzymes, Madyastha et al., (1985) tested the effects of pulegone on cytochrome P450 and found that pulegone when fed to rats at a dose of 400mg/kg once daily for 5 days destroyed cytochrome P450 and haem but the reaction was reversible since discontinuation of treatment returned the enzyme levels to control levels within 4 days. This effect was restricted to pulegone alone since there was no significant loss in cytochrome P450 when the rats were pretreated with menthone or carvone. Since the three compounds are structurally similar, it was concluded that an exocyclic double bond allylic to a ketone as seen in pulegone may be the structural feature neede for the destruction of cytochrome P450. Ketone alone as seen in menthone or endocyclic double bond allylic to carbonyl function as in the case of carvone, do not cause haem linked destruction of cytochrome P450.

The reduction in the level of liver microsomal cytochrome P450 and haem was accompanied by massive hepatotoxicity with an increased serum glutamate pyruvate transaminase and a decreased glucose-6-phosphatase ( Moorthy et al., 1989). A significant decrease in amino pyrine N-demethylase was also noticed after pulegone administration. Pretreatment of rats with phenobarbital potentiated the hepatotoxicity caused by pulegone, whereas, pretreatment with 3-methylcholanthrene protected from it. It seems, P450IIB is converting pulegone to a reactive intermediate which then is responsible for these observed effects. Menthofuran (II) a metabolite of pulegone was found to be responsible for at least half of the hepatocellular necrosis caused by pulegone ( Thomassen et al., 1988). Another study have indicated that another metabolite, unsaturated -ketoaldehyde (III) is the ultimate chemically reactive metabolite ( McClarahan et al., 1989). When menthofuran (II) was incubated with rat liver microsomes in the presence of NADPH and O2, the -ketoaldehyde (III) was formed (Madyastha and Raj, 1990). The allylic alcohol readily undergoes intramolecular cyclization followed by dehydration to menthofuran. It was proposed that the -ketoaldehyde (III) formed from menthofuran (II) contributed to the toxicity mediated by pulegone in two ways:

  1. the -ketoaldehyde can covalently bind to the macromolecules resulting in toxicity.
  2. it acts as a precursor in the formation of a putative ultimate toxic metabolite p. cresol (VIII) (Madyastha and Raj, 1991.


Limonene is a hydrocarbon monoterpene. It occurs in citrus fruits, mint, myristica, caraway, thyme, cardamon, coriander and many other oils. d-Limonene, a major component in many essential oils of citrus fruits is rapidly absorbed from the gastrointestinal tract and rapidly eliminated with no significant accumulation in the rat (Igimi et al., 1974). It is converted in man and in some other species of animals into two glycols, one of which is d-limonene 8,9-glycol as the major metabolite, and the other d-limonene 1,2-glycol as the minor one (Watabe et al., 1981).

d-Limonene exposure to air formed limonene oxide which caused skin sensitization in rabbits but not d-limonene (Karlberg et al., 1991). Out of many terpenes, d-limonene was found to be one of the most effective agents to solubilize cholesterol (Nishimura, 1972) and has been developed as a preparation to dissolve cholesterol gallstones by infusing into the gall bladder and / or the bile duct. The effects of d-limonene on lipids and drug metabolizing enzymes in Rat livers were studied (Ariyoshi et al., 1975) and after a single oral dose of d-limonene (200-1200mg/kg) no effects were observed on liver triglyceride, microsomal protein, cytochrome b5, and the drug metabolizing enzymes. Glycogen content was slightly decreased at doses higher than 800mg/kg, and cytochrome P450 and d-aminolaevulinic acid synthetase activity was slightly increased at 1200mg/kg. After repeated treatment (400mg/kg/day) for 30 days, the relative liver weight and hepatic phospholipid content were only slightly increased, and liver and serum cholesterol were decreased 4 and 8% respectively of the phospholipid fatty acids, palmitic, linoleic and arachidonic acids were increased and stearic acid was decreased. Aminopyrine demethylase and aniline hydroxylase were increased 26 and 22% respectively, and cytochrome P450 and b5 were likewise increased 31 and 30%. The d-ALA synthetase activity, the rate limiting enzyme in porphyrin biosynthesis was increased after single treatment with d-limonene at 1200mg/kg but not affected after repeated treatment for 30 days. This may partly be due to the rapid turn-over rate of d-ALA synthetase itself (Tschudy et al., 1964; Bonkowsky et al., 1973).

d-Limonene was found to have chemopreventive and chemotherapeutic activity against chemically induced mammary ( Elegbede et al., 1984; Maltman et al., 1989), lung and stomach (Wattenberg and Coccia, 1991) cancer in rodents. When fed during the promotion/progression stage of mammary carcinogenesis, limonene inhibited the development of tumours induced by either 7,12 dimethylethylbenz(a) anthracene, which requires metabolic activation to its carcinogenic form or nitrosomethylurea, a directly acting carcinogen. Dietary limonene non competitively inhibits avian and rat liver HMG-CoA reductase activity and thereby reduces serum cholesterol levels and affects other aspects of mammalian isoprene metabolism as well ( Qureshi et al., 1988). In mammals, mevalonic acid is metabolised to many isoprene species i.e.haem, cholesterol, tRNA's, ubiquinones etc. Cells treated with lovastatin (a potent inhibitor of HMG-CoA reductase) cease to grow and do not completely resume growth when cholesterol, ubiquinone or dolichol is added to the medium ( Goldstein and Brown, 1990; Maltese, 1990). Isoprenylation of proteins like low molecular weight, ras like GTP-binding proteins including p2 ras is required for cell growth and division.

Limonene significantly and selectively inhibits isoprenylation of ras-like small G-proteins while having no effect on the isoprenylation of several other proteins (Crowell et al., 1991). The mevalonate pathway produces isoprenoids that are vital for diverse cellular functions, ranging from cholesterol synthesis to growth control. Several mechanisms for feedback regulation of low-density-lipoprotein receptors and two enzymes involved in mevalonate biosynthesis ensure the production of sufficient mevalonate for several end-products. Manipulation of this regulatory system could be useful in treating certain forms of cancer as well as heart disease. The bulk product of mevalonate metabolism, cholesterol is obtained from two sources;

  1. endogenously by synthesis from acetyl-CoA through mevalonate and
  2. exogenously from receptor-mediated uptake of plasma LDL (low-density- lipoprotein).
Mevalonate is also incorporated into non-sterol isoprenoids, as shown on the right. Mevalonate homeostasis is achieved through
  1. sterol-mediated feedback repression of the genes for HMG-CoA synthase, HMG-CoA reductase and the LDL receptor, as shown on the left and
  2. post-transcriptional regulation of HMG-CoA reductase by one of the nonsterol isoprenoids shown on the right.
Cholesterol ---> involved in membrane structure,
Haem A and ubiquinone ---> which partake in electron transport,
Dolichol ---> required for glycoprotein synthesis and
Isopentyladenine ---> present in some transfer RNAs.

Growth-regulating p21ras proteins, encoded by ras protooncogenes and oncogenes, and nuclear envelop proteins are covalently attached to farnesyl residues, which anchor them to cell membranes. Inhibition of mevalonate synthesis prevents farnesylation of these proteins and blocks cell growth. It is possible that the chemopreventive and chemotherapeutic action of d-limonene against mammary, lung and stomach cancers may be due to its inhibition of the mevalonate pathway. There is no information as to if other terpenoids espercially the hydrocarbon monoterpenes similar in structure to limonene (even l-limonene) inhibit mevalonate synthesis in mammals. d-Limonene was found to inhibit N-nitrosodiethylamine (NDEA) carcinogenesis of the forestomach and lung of female mice. d-Limonene and the citrus fruit oils can inhibit the tobacco-specific carcinogen, 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone, NKK-induced carcinogenesis when administered 1 hour prior to carcinogen challenge ( Watterberg and Coccia, 1991). Although limonene can be consumed by rats (Elson et al., 1988; Kodama et al,. 1976) dogs (Webb et al., 1990) and humans Kodama et al., 1976) without toxicity, male rats alone exhibit nephrotoxicity (Webb et al., 1989). Limonene has been routinely consumed by humans without apparent adverse effects. Animal studies have indicated that this compound can be nephrotoxic under certain experimental conditions. Several studies have reported that the mature male rat treated with d-limonene and a wide variety of hydrocarbons, is the only species and sex to develop a unique form of nephrotoxicity that is characterized by the following specific triad of changes observed under the light microscope;

In order to assess if this type of nephrotoxicity occurs in a non-rodent species, the dog was used as a model (Webb et al., 1990), but was found to be refractory to the hyaline droplet nephropathy observed in male rats. Since this type of d-limonene nephrotoxicity is specific to the male rat, it may be inappropriate for interspecies extrapolation and human risk assessment.

The cause of this toxicity was attributed to the reversible binding of d-limonene-1,2-oxide, a metabolite of d-limonene to alpha-globulin (Lehman-McKeman et al., 1989; Dietrich and Swenberg, 1991) a naturally occurring low molecular weight urinary protein that is found in abundance only in adult male rats (Swenberg et al., 1989). The complex formed is more resistant to lysosomal degradation than alpha-globulin alone. The reduced degradation of alpha-globulin-limonene complex leads to an accumulation of this complex in the proximal convoluted tubule leading to the morphological changes characteristic of alpha-globulin nephropathy. Because alpha is a low molecular weight protein it is readily filtered at the glomerulus. Approximately 50% of alpha is absorbed in the proximal tubule and the remainder is excreted in the urine ( Neuhaus et al., 1981).

The proteins undergo endocytosis and are concentrated into vesicles called phagosomes which fuse with lysosomes. Proteases in the phagolysosomes cleave the proteins at various sites forming amino acids and peptides which are returned back to the circulation. The function of the alpha is mainly as a carrier of lipophilic molecules. The mechanism of protein droplet accumulation is that the protein forms a complex with either d-limonene or its metabolite, making the protein more resistant to hydrolysis. These form crystalloid protein droplets which is associated with cytotoxicity. The injured cell is released from the basement membrane and sloughs into the lumen where it collects in granular casts in the thin loop of Henle or is excreted in the urine. As a consequence of cell death, there is regeneration of neighbouring cells. The increased cell proliferation enhances the likelihood of an increase in spontaneous mutational events and also encourages clonal expansion of initiated cells.

The development of alpha-nephropathy is dependent on the presence of alpha, since strains of rats, like the NBR (NCl-Black-Reiter) rat which are deficient in alpha-globulin, female rats and other species that do not synthesize this protein do not develop this disease ( Borghoff, 1990; Dietrich and Swenberg, 1991). Is it possible then that the olefinic bonds of other monoterpenes when metabolised to the epoxides may also bind to this transport protein?