Keywords

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

1 Dietary Phytochemicals and Health Benefits: The Epidemiological Evidence

Many epidemiological studies have sought to quantify the association between diets that are rich in fruits and vegetables with reduced risk of chronic illness such as cardiovascular disease, cancer and cognitive decline. The majority of studies undertaken in the 1980s–2000 were retrospective case control studies, and many of these reported inverse associations between diets rich in fruits and vegetables in general, or particular classes of secondary metabolites that occur in these foods and risk of chronic disease. Case control studies suffer from two major problems. Firstly, they do not give any indication of absolute risk, and thus can give a misleading impression of health benefits. Secondly, selection of the control group is complex and challenging, particularly as diets rich in fruits and vegetables are often highly correlated with other lifestyle attributes and socioeconomic status. Furthermore, due to the large number of studies there is likely to be bias in the reporting of studies that have a positive outcome, as opposed to those that find no association between diet and health. More recently, several long term prospective studies involving many thousands of volunteers have provided further evidence for the associations between diet and health. In general, and with acknowledging that there are several exceptions, longer term prospective studies and pooled analyses of smaller cohort studies have supported the association between diets rich in fruits and vegetables and a reduction in risk of cardiovascular disease [1, 2] and, although there are fewer studies, a reduction in risk of cognitive decline and neurological disorders [3] and chronic obstructive pulmonary disease [4]. Concurrent with the reporting of results from prospective studies has been the establishment of databases of the phytochemical content of fruits and vegetables and other food stuffs, particularly with regard flavonoids [5–7]. Integrating these two datasets has provided evidence for the association of flavonoids, particularly anthocyanidins and proanthocyanidins, with reduction in vascular disease [2]. However, these larger studies have provided less supporting evidence for an association between diets rich in fruit and vegetables and a reduction in cancer risk [8–13], with some notable exceptions. For example, diets rich in cruciferous vegetables, and therefore rich in glucosinolates, have been associated with a reduction in risk of the development of aggressive prostate cancer [14], diets rich in soy, and therefore rich in isoflavones, have been associated with reduction in risk of localised prostate cancer [15], and diets rich in vegetables in general have been associated with reduction in risk of gastric cancer [16].

2 Dietary Phytochemicals and Health Benefits: Evidence from Cell and Animal Studies

Cell and animal models have been used to provide further experimental evidence for the association between phytochemicals and reduction in risk of chronic disease. These models have largely been developed for toxicological research and caution needs to be exercised in extrapolation of results from these models to humans consuming a normal diet. For example, many cell lines that are routinely used have been in culture for several years and were originally derived from cancerous tissues. Both of these factors result in differences in expression of genes involved in, for example, cell proliferation compared to ‘normal’ cells. This is of particular importance if the main interest is in disease prevention rather than cure. Secondly, there is often insufficient appreciation of the bio-transformations and metabolism that dietary phytochemicals undergo within the gastrointestinal tract prior to absorption, either due to mammalian metabolism or colonic fermentation, and, following absorption, within epithelial cells and hepatic tissue. Thus, frequently, cells are exposed to plant metabolites that would not be found within the GI tract or in the systemic circulation. Thirdly, the concentration of dietary phytochemicals to which cells and animal models are exposed is frequently significantly greater (sometimes by a few orders of magnitude) than that which would result from normal dietary exposure. Fourthly, cells and rodent models may have different functional alleles and lack the equivalent allelic polymorphisms that mediate the interaction between phytochemicals and, for example, gene expression. Lastly, even a single item of food consists of a complex mixture of phytochemicals, which may interact, possibly synergistically, with each other.

Despite these reservations, cell and animal models are effective in identifying particular bioactive phytochemicals that are good candidates to underpin the health benefits of fruit and vegetables, even if the nature of the mechanistic basis is questionable.

3 Dietary Phytochemicals and Health Benefits: Evidence from Human Intervention Studies

Obtaining experimental evidence for the health benefit of plant natural products from human intervention studies is challenging. Classic placebo controlled doubled blinded designs, as used for clinical trials of pharmaceuticals, are not possible except for the use of dietary supplements, in which specific phytochemicals are provided at relatively high doses. Likewise, most of the evidence for the protective effect of plant natural products is towards forms of chronic diseases that are associated with aging. These may take many years to develop and preclude the use of clinical endpoints in intervention trials. Thus, researchers are forced to evaluate ‘biomarkers’ of risk. While these are well defined for vascular disease, there are few biomarkers appropriate for evaluating cognitive decline and cancer risk.

The positive results of associations obtained from case control studies supported by mechanistic studies with cell models led to the establishment of intervention trials with dietary supplements. However, these have frequently either not resulted in supporting the association between phytochemicals and health, at least with the dose that were provided which tended to be higher than dietary intake, or suggested that the supplements may increase risk. For example, while certain epidemiological studies with carotenoid rich foods have suggested that β-carotene and vitamins A and E may have protective effects against lung cancer, intervention trials have found either no protective effect or an increase in cancer risk [17, 18]. In a similar manner, in a prospective epidemiological study (i.e. not an intervention study) it was found that men who were heavy users of multivitamin supplements had enhanced risk of advanced and fatal prostate cancer [19].

In contrast, many relatively short intervention trials with either supplements or foods have provided supporting evidence for the benefits of flavonoids for vascular health. The majority of these studies are of an acute nature in which relatively high doses are provided and various biomarkers, including, for example, platelet aggregation [20], flow mediated arterial dilation and nitric oxide metabolism [21], and plasma LDL cholesterol [22] are quantified over the following hours or days. Some studies have also demonstrated increased cerebral blood flow and cognitive function [23]. The major challenge is to design intervention studies in which health benefits of normal dietary intake can be assessed over several months or possibly a few years, and to go beyond the assessment of ‘biomarkers’ to probe the underlying mechanisms in vivo.

4 Dietary Phenolics, Polyphenolics, Tannins: Structure and Human Metabolism

The chemistry and biochemistry of flavonoids and related compounds has recently been extensively reviewed [24]. Over 8,000 compounds with phenolic structures have been reported in plants, ranging from low molecular weight simple phenolic acids to high molecular weight tannins. They can be conveniently divided into two major groups, the flavonoids, which are compounds comprising two aromatic rings connected by a three carbon bridge, and the non flavonoids, including simple phenolic acids, hydroxycinammates and stilbenes. Flavonoids can further be divided into six major subclasses: the flavanols, flavan-3-ols (comprising both monomers and the polymeric proanthocyanidins or condensed tannins), isoflavones, flavones, flavanones and anthocyanidins (Figs. 1 and 2). The principle phenolic acid is gallic acid which is the base unit of gallotannins, whereas gallic acid and hexahydroxydiphenoyl are subunits of ellagitannins. Both forms of tannins are collectively known as hydrolysable tannins. The most common hydroxycinna-mates are caffeic, p-coumaric and ferulic acids. Quinic acid conjugates of caffeic acid such as 5-O-caffeoylquinic acid (chlorogenic acid) are common components of fruit and vegetables, with coffee being a major source in the diet. Resveratrol, an important polyphenolic found in red wine, is the most common stillbene, and has frequently been associated with health benefits (Fig. 3).

Fig. 1
figure 1_18

The basic C6-C3-C6 flavonoid skeleton and examples of common glycosylated flavonoids that are found in foods

Full size image
Fig. 2
figure 2_18

Examples of flavan-3-ols

Full size image
Fig. 3
figure 3_18

Examples of common non flavonoid phenolics that are found in foods and are associated with health benefits

Full size image

While there is extensive knowledge of the structure and distribution of these compounds in plants, and a growing understanding of their biosynthesis, there is a relatively poor understanding of the bioactive functional derivatives of these compounds that occur either within the gastrointestinal tract or in the systemic circulation following consumption. This has often led to the use of inappropriate compounds within model systems that have attempted to elucidate function, as discussed above. Thus, prior to discussing potential modes of action, the biochemical processes involved in absorption, biotransformation and metabolism will be briefly reviewed.

Almost all flavonoids with the exception of the flavan-3-ols are glycosylated, and hydrolysis of the glycoside is a prerequisite for absorption. Deglycosylation can occur by a variety of routes. Flavanoid glycosides such as quercetin glucoside are hydrolysed by the endogenous β-glucosidases lactase phloridzin hydrolyase (LPH) present on the brush border of small intestine epithelial cells [25]. The resulting aglycone can passively diffuse into the epithelial cells. Alternatively, the glycosylated flavonoids may be actively transported into epithelial cells by, for example, the sodium-dependent glucose transorter (SGLT1) [26], and deglycosylated through the activity of broad specificity cytosolic β-glucosides (CBG) [27]. The relative importance of these two routes is likely to depend upon the position and extent of glycosylation, which determines both the efficiency of transport into the cells by glucose transporters and subsequent specificity of the glucosides. Glycosylated dietary flavonoids may compete with glucose for transport into cells via SGLT1, and this may have broader biological importance in modulating glucose transport. Some studies have suggested that polyphenol rich diets may reduce the post prandial surge in plasma glucose, and thus effectively reduce the glycaemix index of foods consumed as part of a polyphenol rich diet [28]. This may in itself have potential health benefits. Some flavanoid glycosides, such a quercetin rhamnoglucosides (rutin) cannot act as substrates for either LPH or CBG and are deglycosylated by microbial glucosidases in the colon, prior to absorption.

Non glycosylated flavanoids such as epigallocatechin gallate and other phenolics such as 5-O-caffeoylquinic acid (chlorogenic acid) may also be metabolised in the gastrointestinal tract prior to absorption. Hydrolysis of epigallocatechin gallate has been suggested to be due to an esterase occurring in human saliva, whereupon the main site of hydrolysis of 5-O-caffeoylquinic acid is likely to be in the colon due to microbial activity.

The role and importance of the microbial metabolism and transformation of phenolics and polyphenolics by the microflora of the colon, or chemical transformation within the GI tract or systemic circulation has probably been underestimated. While it is increasingly likely that oligomeric and polymeric proanthocyanidins are extensively metabolised in the colon to produce a range of small phenolic acids that are absorbed [29], other compounds such as anthocyanidins are also likely to be metabolised in an analogous manner. Indeed it is possible that these cleavage products are the most relevant biologically active forms of dietary phenolics in vivo. For example, while several studies have reported that less that 1% of anthocyanins that are consumed are absorbed and excreted, which is difficult to reconcile with the reported health benefits of these compounds, an intervention study with blood orange juice reported that 44% of ingested cyanidin glycoside was accounted for in the plasma in the form of protocatechuic acid, a cleavage product of anthocyanidins (Fig. 4) [30, 31]. Elucidating these metabolic processes and identifying the precise metabolites that are absorbed are an important prerequisite for understanding biological function and the mechanistic basis of the health benefits of phenolics and polyphenolics.

Fig. 4
figure 4_18

Cleavage products from anthocyanins (Redrawn from Ref. [31])

Full size image

Once absorbed, flavanoids and phenolic acid derivatives undergo extensive metabolism. Initially extensive glucuronidation and some methylation of the aglycone occur in the cells of the small intestine, with further glucuronidation, methylation and sulphation occurring in the liver. Thus, following ingestion and deglycosylation a single polyphenolic may give rise to several conjugated metabolites. For example, following ingestion of quercetin glycoside, at least 12 different glucuronide and sulphate conjugates of quercetin or methylquercetin are found in plasma (Fig. 5) [32, 33]. Likewise, several glucuronide and sulphate conjugates are observed following ingestion of other polyphenols, such as the isoflavones [34], and it is likely that similar compounds are also derived from other flavonoids. Non flavanoid phenolics follow similar routes of absorption and metabolism. For example, hydroxycinnamic acids such as caffeic, ferulic and coumaric acids, are rapidly absorbed in the small intestine and glucuronidated and sulphated in a similar manner to the flavanoids [35].

Fig. 5
figure 5_18

The most common conjugates of quercetin that are found in human plasma

Full size image

5 Mechanistic Basis of Health Promoting Activity of Dietary Phenolics and Polyphenolics

5.1 Polyphenols as Antioxidants

The antioxidant activity of phenolics and polyphenolics is often regarded to be the basis of their health promoting activity. The evidence for this is largely based upon cell culture studies with aglycones, as opposed to the more appropriate conjugated metabolites that occur in vivo. Several studies have also attempted to demonstrated that a polyphenol rich diet results in an increase in plasma antioxidant capacity, but few, if any, have demonstrated a significant increase [36]. This is due to two main factors: Firstly, the endogenous phenolic and ascorbate concentrations in the plasma is between 159 and 380 μM. The additional concentrations that can be obtained from dietary sources is relatively low, probably less that 1% from average diets and up to 5% for diets that are particularly rich in polyhenols [36]. Moreover, these additional marginal increases are transient and it is difficult to envisage how these changes can have a significant impact upon health. However, it is conceivable that in certain elderly populations in which the plasma ascorbate levels can become depleted heavy consumption of tea and coffee may have a significant effect on plasma antioxidant activity. Secondly, the conjugated metabolites of the aglycones often have reduced antioxidant activity compared to the parent agylcone, with the precise activity dependant upon the nature and position of conjugation [37, 38]. For example, sulphation has been shown to reduce the antioxidant activity of isoflavones [39]. In addition, the interaction between polyphenols and plasma proteins can reduce their antioxidant activity [40].

Counter to these arguments, there are two factors that may lead to the antioxidant activity of human metabolites of dietary polyphenols being underestimated. Firstly, as described above, there is some uncertainty to the precise metabolic derivatives of polyphenols and thus the plasma concentration of the active metabolites may be underestimated. Secondly, it is possible that there may be local deglucuronidation at, for example, sites of inflammation through β-glucuronidase activity to release the biological active aglycone [41]. However, despite the possible factors in mitigation, while the epidemiological evidence for health promoting activity of polyphenolic rich diets has increased, it is unlikely that this is mediated by enhancement of antioxidant activity of plasma.

5.2 Polyphenols and Vascular Disease

Partially due to the epidemiological evidence that has associated diets rich in polyhenols with reduction in risk of cardiovascular disease, several studies have specifically investigated the effects of polyphenolics on risk factors for cardiovascular health. Inflammation plays a key event in the initiation of atherosclerosis and the development of atherothrombotic events, which are leading causes for CVD. Adhesion of circulating monocytes to the endothelium and subsequent migration into the vascular wall are critical events in these processes. The binding of monocytes to the vascular endothelium is mediated by cross linkage of cell adhesion molecules, such as intercellular adhesion molecule 1 (ICAM-1) and vascular cell adhesion molecule-1 (VCAM-1), and the expression of these molecules is greatly enhanced at sites of atherosclerosis. The expression of these adhesion molecules is enhanced through several risk factors for CVD, such as increased plasma cholesterol and pro-inflammatory cytokines, such as TNF-α and IL-1, mediated through activation of the NF-κB and AP-1 transcription factors. Through the use of HUVEC primary cells, it has been shown that quercetin conjugates at physiologically appropriate concentrations inhibit cell surface expression of VCAM-1, and that quercetin 3-glucuronide, but not other conjugates, also inhibit ICAM-1 expression. Interestingly, these effects were not observed at higher concentrations [42]. The precise mechanism involved in inhibition of cell adhesion molecules is not understood, but may involve inhibition of NF-κB-mediated signalling. Several other studies with different cell lines and polyphenols, including flavanones, isoflavones, anthocyanidins and catechins, are consistent with these results, although many of these use only the aglycones, as opposed the physiologically appropriate meta-bolites [43, 44]. Moreover, it has been shown that red wine consumption in humans reduces TNF-α induced adhesion of monocytes to endothelial cellsex vivo [45].

In addition to changes in expression of adhesion molecules, polyphenols have been shown to affect other risk factors for CVD. For example, several polyphenols, including quercetin, epigallocatechin gallate and resveratrol have been shown to inhibit vascular smooth muscle cell proliferation [46–49] and can modulate the response of platelets to thrombin and other agents [50–52].

5.3 Polyphenols and Plasma Protein Interactions

The oral sensation of dryness known as astringency is probably the most familiar example of a polyphenol–protein interaction. Salivary proteins contain multiple binding sites for polyphenols, and when sufficiently high concentrations of polyphenols, and in particular tannins, are present the polyphenol–protein complexes precipitate leading to the taste sensation [53]. Several polyphenols are also known to inactivate digestive enzymes in the gut [54]. Poly-phenols can bind with plasma proteins by hydrophobic interactions, hydrogen bonds, and covalent bonds. The nature of the interaction depends upon the structure of the polyphenol, so that there may be a combination of non specific binding, and binding to specific proteins. It has been shown that quercetin will bind with plasma albumins [55], epigallocatechin-gallate binds with plasma fibronectin and fibrinogen [56], and wine catechins with Apo-A1 and transferrin [57]. It is likely that these interactions are far more extensive, and while they have mainly been considered in the context of polyphenol transport in plasma, they may themselves mediate the biological activity of polyphenol metabolites, either through effects such as reducing or enhancing antioxidant activity [40, 57], or interaction with, for example, pro-inflammatory cytokines.

Interactions with ligand receptor proteins have largely been restricted to the isoflavones which are structural mimics of estrogens [58]. Estrogen hormones influence the growth and functioning of many tissues of the male and female reproductive systems. Isoflavones, and in particular genistein, are structural mimics of estrogens and can tightly bind to the estrogen receptors α and β, and act as estrogen agonists [59, 60]. This phenomenon has been offered for an explanation for the possible protective effects of diets rich in soy towards breast and prostate cancer, but it is likely other mechanisms are also of importance.

6 Glucosinolates and Isothiocyanates: Structure and Human Metabolism

Epidemiological studies have suggested that diets rich in cruciferous vegetables, such as broccoli, may reduce the risk of cancer and myocardial infarction. It is widely thought that isothiocyanates, derived from glucosinolates that accumulate in cruciferous vegetables, are the active component although indole degradation products from tryptophan-derived glucosinolates and other polyphenolic compounds, as discussed above, that are found in these vegetables may also play a role. The glucosinolate molecule consists of a β-thioglucose moiety, a sulfonated oxime moiety and a variable side chain, derived from an amino acid. Glucosinolates with more than 120 side chain structures have been described [61], although only about 16 of these are commonly found within crop plants. Seven of these 120 side chain structures correspond directly to a protein amino acid (alanine, valine, leucine, isoleucine, phenylalanine, tyrosine and tryptophan). The remaining glucosinolates have side chain structures which arise in three ways. Firstly, many glucosinolates are derived from chain-elongated forms of protein amino acids, notably from methionine, but also from phenylalanine and branch chain amino acids. Secondly, the structure of the side chain may be modified after amino acid elongation and glucosinolate biosynthesis by, for example, the oxidation of the methionine sulfur to sulfinyl and sulfonyl, and by the subsequent loss of the ω-methylsulfinyl group to produce a terminal double bond. Sub-sequent modifications may also involve hydro-xylation and methoxylation of the side chain. Chain elongation and modification interact to result in several homologous series of glucosinolates, such as those with methylthioalkyl side chains ranging from CH3S (CH2)3- to CH3(CH2)8-, and methylsulfinylalkyl side chains ranging from CH3SO(CH2)3- to CH3SO(CH2)11-. Thirdly, some glucosinolates occur which contain relatively complex side chains such as o-(α-L-rhamnopylransoyloxy)-benzyl glucosinolate in Reseda odorata and glucosinolates containing a sinapoyl moiety in Raphanus sativus. Several comprehensive reviews of glucosinolate structure and biosynthesis have recently been published [62–66].

Despite the potential large number of glucosinolates, the major cruciferous crops have a restricted range of glucosinolates. All of these have a mixture of indolylmethyl and N-methoxyindolylmethyl glucosinolates, derived from tryptophan, and either a small number of methionine-derived or phenylalanine-derived glucosinolates. The greatest diversity within a species is found in B. oleracea, which includes such crops as broccoli, cabbages, Brussels sprouts and kales. These contain indolyl glucosinolates combined with a small number of methionine-derived glucosinolates. For example, broccoli (B. oleracea var. italica) accumulates 3-methylsuphinylpropyl and 4-methylsulfinylbutyl glucosinolates, while other botanical forms of B. oleracea have mixtures of 2-propenyl, 3-butenyl and 2-hydroxy-3-butenyl. Some cultivars of cabbage and Brussels sprouts also contain significant amounts of methylthiopropyl and methylthiobutyl glucosinolates. B. rapa (Chinese cabbage, Bok Choi, turnips etc.) and B. napus (Swedes) contains 3-butenyl and sometimes 4-pentenyl glucosinolates and often their hydroxylated homologues. In addition to methionine-derived glucosinolates, phenylethyl glucosinolate usually occurs in low levels in many vegetables. Several surveys of glucosinolate variation between cultivars of Brassica species have been reported, for example B. rapa [67, 68] and B. oleracea [69, 70].

The distinctive taste of many minor horticultural cruciferous crops is due to their glucosinolate content. For example, watercress accumulates large amounts of phenylethyl glucosinolate, combined with low levels of 7-methylsulfinylheptyl and 8-methylsulfinyloctyl glucosinolates, rockets (Eruca and Diplotaxis species) possess 4-methylthiobutyl glucosinolate, and cress (Lepidium spp) contains benzyl glucosinolate.

In the intact plant, glucosinolates are probably located in the vacuole of many cells but may also be concentrated within specialized cells. Following tissue disruption, glucosinolates are hydrolysed by thioglucosidases, known as myrosinases (Fig. 6). Myrosinase activity results in the cleavage of the thio-glucose bond to give rise to unstable thiohydroximate-O-sulfonate. This aglycone spontaneously rearranges to produce several products [71]. Most frequently, it undergoes a Lossen rearrangement to produce an isothiocyanate (ITC). Aglycones from glucosinolates which contain β-hydroxylated side chains, such as 2-hydroxy-3-butenyl (‘progoitrin’) found in the seeds of oilseed rape and some horticultural brassicas, such as Brussels sprouts and Chinese cabbage, spontaneously cyclise to form the corresponding oxazolidine-2-thiones. If the isothiocyanate contains a double bond, and in the presence of an epithiospecifier protein (ESP), the isothiocyanate may rearrange to produce an epithionitrile [72, 73]. ESP is also likely to be involved in the production of nitriles from glucosinolates such as methylsulfinylalkyls [74]. Cooking can denature both ESP and myrosinase [75], in which case intact glucosinolates can be metabolized by microbial thioglucosidases in the colon to generate isothiocyanates [76].

Fig. 6
figure 6_18

Glucosinolate structure and metabolism, and side chains (R) occurring in common cruciferous vegetables

Full size image

Following ingestion of cruciferous vegetables with intact myrosinase, ITCs will be formed in the mouth and rapidly absorbed in the upper GI and subsequently metabolized[77, 78]. When myrosinases in the plant tissue are deactivated, e.g. by excess cooking, then glucosino-lates are hydrolysed in the distal gut by microbial activity, and the resulting ITCs are absorbed from the lower GI tract [78]. Conjugation with glutathione occurs spontaneously but may be further promoted by gluthatione transferases (GST) within the epithelial cells of the GI tract. The glutathione conjugate is then exported to the systemic circulation via the multidrug resistance associated protein-1 (MRP1), MRP2 and P-glycoprotein-1 (Pgp-1) [79, 80].

The ITC-glutathione conjugate is metabolised via the mercapturic acid pathway in which the glutathione conjugate undergoes further enzymatic modifications including cleavage of glutamine, which yields cysteine-glycine- conjugates, cleavage of glycine, yielding cysteine-conjugates and finally acetylation to produce N-acetylcysteine (NAC)-conjugates that are excreted in urine (Fig. 6) [81]. However, it has been shown that 45% of ingested sulforaphane (SF, the major isothiocyanate derived from 4-methylsulphinylbutyl glucosinolate that accumulates in broccoli) in the plasma occurs as the free ITC, as opposed to thiol conjugates, and it has been speculated that the ITC-glutathione conjugate may be cleaved in the plasma to release the free, and biologically active ITC, possibly through GSTM1 activity [77]. The peak concentration of SF and its thiol conjugates following consumption of a standard portion of broccoli is less than 2 μM, falling to low (nM) levels within a few hours [77].

7 Diet-Gene Interactions, and the Role of GSTM1 Genotype

Epidemiological evidence both from prospective cohort studies and retrospective case-control studies suggest that there is an inverse association between consumption of cruciferous vegetables and the risk of lung, stomach, colorectal, breast and prostate cancer [14, 82–91]. Several of these studies suggest that the protective effects of crucifer consumption are modulated by GST polymorphisms, and in particular GSTM1 genotype. Fifty percent of the population have a homozygous deletion of the GSTM1 allele. Studies of US populations have suggested that individuals with GSTM1 positive genotype benefit more from consumption of brassica vegetables compared to those that have a homozygous GSTM1 deletion (i.e. GSTM1 null) [82, 85, 91, 92]. However, similar studies conducted on Asian populations have found the converse; with GSTM1 nulls gaining greatest benefit [87, 88 ]. It has been speculated that this may be due to the contrasting types of vegetables being consumed; broccoli is the major crucifer consumed in the US, whereas in Asia the major cruciferous vegetable consumed is Chinese cabbage. These two vegetables have contrasting types of glucosinolates which may interact with GSTM1 in different ways, and as previously described for structural similar ITCs [93]. Polymorphism also occurs at the GSTT1 locus, with about 20% of Caucasians and up to 60% of people of Asiatic descent having a homozygous deletion. GSTT1 genotype has been associated with modulating the reduction in cancer risk through cruciferous vegetable consumption in some studies, but not others [85, 94, 95]. Although GSTT1 has not been extensively studied there is some evidence that a combination of GSTM1 and GSTT1 genotype might influence cancer risk following brassica consumption [87]. Recently, consumption of cruciferous vegetables was also associated with a lower risk of myocardial infarction among those individuals with a functional GSTT1 allele [96].

Complementary to the epidemiological data, a few experimental studies have shown that GSTM1 null genotypes excrete a greater proportion of ingested SF via mercaturic acid metabolism than those with at least one functional allele [77, 97]. As it is the latter who gain more protection this suggests that there may be other metabolic fates for ITC. Further studies are required.

8 Mechanistic Basis of Health Promoting Activity of Dietary Glucosinolates and their Derivatives

The anticarcinogenic activity of cruciferous vegetables has largely been attributed to the biological activity of isothiocyanates, although degradation products from indole glucosinolates may also play a role. The activity of sulforaphane, the ITC derived from broccoli has recently been comprehensively reviewed [98]. Other ITCs have similar activity. Sulforaphane has been shown to be protective against carcinogen-induced tumorigenesis at a variety of sites in rodents, including breast, colon, skin, lung, stomach and prostate. It is effective in reducing and even preventing the formation of preneoplastic lesions in tissues resulting from carcinogen administration [99–102] and can also suppress the growth of tumours in spontaneous or xenograft mouse cancer models [103, 104 ]. The chemopreventive effect of SF is likely to involve multiple mechanisms, which are likely to interact together to reduce risk of carcinogenesis [98]. These include: inhibition of phase 1 enzymes, induction of phase 2 metabolism enzymes, antioxidant functions through increased tissue GSH levels, apoptosis-inducing properties, induction of cell cycle arrest, anti-inflammatory properties and inhibition of angiogenesis. As discussed above, certain caution is required in interpreting these results as these mechanistic studies are usually undertaken with far higher concentrations of ITCs than that which would occur following normal dietary consumption of cruciferous vegetables.

During phase 1 metabolism molecules, including dietary and environmental carcinogens, are converted into highly reactive intermediates that can potentially be harmful by binding to critical macromolecules such as DNA, RNA and protein. SF potently decreased enzyme activities of several cytochrome P450 enzymes (CYPs), which catalyse phase 1 biotransformation, in intact human and rat hepatocytes [105, 106]. Activated carcinogens generated from phase 1 metabolism are subsequently converted into inactive metabolites during the phase 2 metabolism and can readily be excreted from the body. SF has received much attention over the past decade as it was found to be the most potent naturally-occurring inducer of phase 2 enzymes such as quinone reductase (NQO1), GST and genes related to glutathione biosynthesis in both animals and humans [107–110]. Induction of phase 2 enzymes is mediated by the nrf2/Keap1 pathway and exposure of cells to SF leads to dissociation of the Nrf2/Keap1 complex and subsequent nuclear translocation of Nrf2 where it activates cancer protective genes [111].

There are a few studies in humans that have sought to provide evidence for the induction of phase II enzymes in vivo following consumption of glucosinolate or isothiocyanate rich diets. A three week diet rich in cruciferous vegetables has been shown to elevate plasma levels of GSTs and reduced the level of 8-oxo-7,8-dihydro-2’-deoxyguanosine in urine, a marker of oxidative damage [112, 113]. However, an intervention study in Qidong region in the People’s Republic of China, where the residents are at high risk of developing hepatocellular carcinoma, partly due to consumption of aflatoxin-contaminated foods, and partly due to exposure to air pollutants such as phenanthrene, did not produce any evidence of differences in excretion of phase 2 metabolic products of aflatoxin and phenanthrene between a diet rich in glucosinolates (via broccoli sprouts) and the placebo control [114]. The large intra-individual variation in excretion of ITC metabolites, which may be partially caused by GST polymorphisms, was noted in this study. A further study quantified global gene expression in gastric mucosa samples after broccoli consumption. While there was evidence for the induction of phase 2 gene expression following consumption of a broccoli that had elevated levels of glucosinolates [115], there was no evidence for induction following consumption of standard broccoli [116]. However, there is evidence that topical application of ITCs to skin can induce phase 2 enzymes [117].

Using cancer cell models ITCs have also been shown to induce apoptosis and cell cycle arrest through a variety of pathways depending on the origin of cells used. The mechanism for such an effect involves induction of several members of the caspase family, responsible for the execution of apoptosis in higher eukaryotes, as well as the induction of the pro-apoptotic Bcl-2 family members. Alternatively, SF can also induce apoptosis through induction of any of the three parallel MAPK cascades identified in mammalian cells, extracellular signal-regulated kinase (ERK), c-Jun N-terminal kinase (JNK) and p38, depending on tissue and dose applied. In line with cancer suppressing properties SF has been shown to arrest the cell cycle through regulation of cyclin levels and induction of p21, a potent inhibitor of cell cycle progression. Additionally, SF down-regulates important mediators of the pro-inflammatory response such as iNOS, Cox-2 and NF-κB, and also inhibits angiogenesis, both processes thought to be mechanistically linked with carcinogenesis.

This brief review has focused on the two classes of plant natural products – flavonoids and glucosinolates – for which the epidemiological evidence for health benefits is strongest. This is, of course, largely due to the frequent occurrences of these compounds in crops and foods. Other natural products that have a limited distribution in certain foods may indeed have far greater health benefits, but these would not be evident from epidemiological studies. The greatest challenge in this field of research is to design human intervention studies of sufficient length coupled with analyses of target tissues to provide a mechanistic understanding of the activity of plant natural products in vivo. The use of genetically modified plants with specific alterations in natural product profiles will greatly facilitate these studies.

9 Glossary

Case control studies

Prospective studies

Flavonoids

Glucosinolates

Antioxidant

Atherosclerosis

Isothiocyanates

Phase 2 metabolism