Polyphenolic Phytochemicals in Cancer Prevention and Therapy

Jun 27, 2017 - Salvador Mena has a Bachelor in Biochemistry (University of Valencia, 2004). a Ph.D. in Physiology with a “Special Award” for molec...
1 downloads 18 Views 5MB Size
Perspective pubs.acs.org/jmc

Polyphenolic Phytochemicals in Cancer Prevention and Therapy: Bioavailability versus Bioefficacy José M. Estrela,*,† Salvador Mena,† Elena Obrador,† María Benlloch,‡ Gloria Castellano,‡ Rosario Salvador,† and Ryan W. Dellinger§ †

Department of Physiology, University of Valencia, 46010 Valencia, Spain Department of Health and Functional Valorization, San Vicente Martir Catholic University, 46008 Valencia, Spain § Elysium Health Inc., New York, New York 10012, United States ‡

ABSTRACT: Natural polyphenols are organic chemicals which contain phenol units in their structures. They show antitumor properties. However, a key problem is their short half-life and low bioavailability under in vivo conditions. Still, definitively demonstrating the human benefits of isolated polyphenolic compounds (alone or in combination) using modern scientific methodology has proved challenging. The most common discrepancy between experimental and clinical observations is the use of nonphysiologically relevant concentrations of polyphenols in mechanistic studies. Thus, it remains highly controversial how applicable underlying mechanisms are with bioavailable concentrations and biological half-life. The present Perspective analyses proposed antitumor mechanisms, in vivo reported antitumor effects, and possible mechanisms that may explain discrepancies between bioavailability and bioefficacy. Polyphenol metabolism and possible toxic side effects are also considered. Our main conclusion emphasizes that these natural molecules (and their chemical derivatives) indeed can be very useful, not only as cancer chemopreventive agents but also in oncotherapy.



INTRODUCTION

polyphenols may influence carcinogenesis as well as antitumor growth and spread,7,8 a limited number of clinical trials have examined the involvement of polyphenols in cancer prevention, incidence, or mortality.9 Reviews and meta-analysis on the possible chemopreventive effects of polyphenols on different cancer types agree that existing evidence does not sufficiently support a significant role of polyphenols in chemoprevention.10 This conclusion is probably linked to (1) impossibility of reproducing in vivo conditions and different host and lifestyle factors, and (2) low level of understanding of the biological activity, toxicity, and (patho)physiological processes in the host.11 However, many of these studies are estimates based on food intake-related questionnaires answered by the participants. At present, if one only considers humans studies (epidemiological or controlled clinical trials) which examined biomarkers of exposure for polyphenols, then only a protective role of isoflavones against breast cancer and prostate cancer among

Cancer (a name given to a collection of related diseases) represents an abnormal and uncontrolled growth of cells that can disrupt normal host functions and, in some cases, spread throughout the body.1 With the continuous rise in life expectancy, in global terms, it is expected that the number of deaths caused by cancer will at least double during the present century.2 Herbal medicine treatments, used for millennia, are common in traditional medicines. Similarly many fruits and vegetables (or products derived therefrom) are consumed for their supposed health benefits.3 Phytochemicals which have been identified and extracted from terrestrial plants for their antitumor properties include, e.g., polyphenols, taxol analogues, vinca alkaloids, or podophyllotoxin analogues.3,4 Epidemiological studies examining a possible relationship between fruit and vegetable consumption and cancer incidence have generated conflicting results. Some studies indicate a positive correlation while others showing no association at all.5,6 Although many experimental results suggest that © 2017 American Chemical Society

Received: July 12, 2016 Published: June 27, 2017 9413

DOI: 10.1021/acs.jmedchem.6b01026 J. Med. Chem. 2017, 60, 9413−9436

Journal of Medicinal Chemistry

Perspective

Figure 1. continued

9414

DOI: 10.1021/acs.jmedchem.6b01026 J. Med. Chem. 2017, 60, 9413−9436

Journal of Medicinal Chemistry

Perspective

Figure 1. continued

9415

DOI: 10.1021/acs.jmedchem.6b01026 J. Med. Chem. 2017, 60, 9413−9436

Journal of Medicinal Chemistry

Perspective

Figure 1. Chemical nature of natural polyphenols and their anticancer effects elicited under in vivo conditions. Polyphenols are classified in nonflavonoids (panels 1 and 2) and flavonoids (panels 3 and 4). Examples of specific compounds and typical natural sources are shown for most subfamilies based on their abundance in nature. Polyphenol metabolites and aglycones circulate in the blood mainly bound to proteins, particularly albumin. Each polyphenolic structure is linked to published work(s) where an antitumor effect has been observed under in vivo conditions, but only if a potential mechanism has been proposed. It is also indicated if that mechanism has an in vivo confirmation. The effects of associations between two different polyphenols are specifically shown in panel 5.7,25−76

Asian populations can be concluded.10 For instance, the Shanghai Breast Cancer Survival Study showed that women who ingested high levels of phytoestrogenic polyphenols had a 29% lower risk of relapse and death.12 A significant number of additional studies are studying the relationship between phytochemicals consumption and cancer, and the largest example underway is the DietCompLyf prospective trial involving more than 3000 women treated for breast cancer in the UK.13 To date, most of all clinical trials involving polyphenols and cancer have included only a limited number of these phytochemicals, i.e., curcumin (59 clinical trials, 20 completed), EGCG (37 clinical trials, 16 completed), genistein (34 clinical trials, 14 completed), resveratrol (20 clinical trials, 8 completed), and quercetin (13 clinical trials, 4 completed).9 In these studies, healthy subjects, patients with a premalignant diagnosis or cancer received orally (or topically, but just in a few cases involving skin cancers) administered polyphenolic phytochemicals. These types of studies often contain mixed chemoprevention- and chemotherapy-related protocols. The observed results show variable results precluding definitive conclusions. Nevertheless, the available evidence supports the advancement of EGCG (or extracts of green tea polyphenols) into phase III clinical trials aimed to prevent progression of prostate cancer, leukoplakia, or premalignant cervical disease. Also, results in phases I and II using curcumin or genistein in different cancers (colorectal and pancreatic cancer for curcumin and bladder cancer, prostate cancer, NSCLC, colorectal, and pancreatic cancer for genistein) support their advancement into phase III trials.14 For example, two phase III clinical trials involving gemcitabine, curcumin, and celebrex in patients with metastatic colon cancer or advanced pancreatic cancer are under development.9 One of the main goals of this Perspective

is to express the need to re-evaluate the basic and clinical research on polyphenols. For example, prospective clinical trials should consider IV administration or, if the oral intake is used, improved delivery systems must reach the level of a coherent correlation between bioavailability and bioefficacy. Natural polyphenols are subject to rapid in vivo metabolism (see below). Thus, as initially proposed by Kroon et al.,15 “in vitro studies on the biological responses to polyphenols should use only physiologically relevant concentrations of these molecules or their conjugates”. This implies the obvious need for establishing mechanisms based on bioavailability. Therefore, a clear need to set strong limitations in terms of (a) real concentrations reached in a specific organ or tissue, and (b) time (cells are usually exposed to biologically active polyphenol concentrations only for a short period of time) is required. Nevertheless, as discussed below, there is enough evidence indicating that some polyphenols show clear antitumor effects in vivo.



THE CHEMICAL NATURE OF POLYPHENOLIC PHYTOCHEMICALS Polyphenolic phytochemicals comprise one of the most ubiquitous class of secondary metabolites found in nature. Chemically, polyphenols are structures with phenyl rings substituted by one or more hydroxy groups. This class contain both simple compounds and much more complex structures, hydrolyzable and condensed tannins for example. This “tannin” term derives from the French word “tan” (powdered oak bark extracts), which is derived from the old Celtic lexical root “tann-” meaning oak.16 Nevertheless it is important to note that only part of the numerous chemical structures identified as polyphenols can be consider as real vegetable tannins, while 9416

DOI: 10.1021/acs.jmedchem.6b01026 J. Med. Chem. 2017, 60, 9413−9436

Journal of Medicinal Chemistry

Perspective

resonance, (2) the presence of OH groups arranged in an ortho configuration in ring B, possibly participating in electron delocalization, (3) the presence of a carbonyl conjugated with one double bond in ring C and forming hydrogen bonds with OH in position 3 (ring C) and position 5 (ring A), (4) the existence of conjugated double bonds (e.g., CC with CC) in ring B provides electron delocalization from ring B, (5) the existence of OH in position 3 on ring A, (6) ether groups being present in rings C and B, and (7) an OH in ring A influencing its antioxidant activity.77 Although basic structures of flavonoids are aglycones, in food, they are mostly β-glycosides. Furthermore, the nature of the glycosylation markedly influences the efficiency of their absorption in the small intestine or in the large intestines after bacterial deconjugation.78 The structural differences and the glycosylation patterns, in combination, dictate the biological activity of these compounds, including their antioxidant activity.20 Indeed, flavonoids have some structural characteristics required for efficient antioxidant activity: (a) an ortho-dihydroxy (catechol) structure on ring B, which confers stability to the radical formed following the capture reaction of a free radical, (b) the 2,3double bond in conjugation with a 4-oxofunction of a carbonyl group in ring C, and (c) the presence of OH groups at the 3 and 5 position in ring A.79,80 With these characteristics in mind, it has been shown that the position of the OH groups, the dipole moment magnitude, and the shape of the molecule play a key role in the inhibition of lipid peroxidation.81 The glycosylation of flavonoids decreases their activity when compared to their aglycones, while alkyl and methoxy groups typically increase their activity. Whereas O-methylation may reflect steric effects that perturb planarity as well as decrease antioxidant activity and hydrophilia. Flavonoids containing a cinnamate or gallate moiety show higher antioxidant activity.82 Moreover, phenolic compounds may also exert antioxidant activity by chelating metal ions.83 However, polyphenols may also behave as prooxidants under certain conditions such as high pH and high concentrations of transition metals in the presence of O2.83,84 Stilbenoids, as examples of nonflavonoid polyphenols, are hydroxylated derivatives of stilbene and have a C6−C2−C6 structure. Stilbenoids are classified phenylpropanoid/polyketide hybrids that share the majority of the chalcone biosynthetic pathway. Subjected to UV radiation, stilbenes (two aromatic rings linked by a ethylene bridge) or their derivatives undergo intramolecular cyclization to form dihydrophenanthrenes.85 Most naturally occurring stilbenes present the E configuration (trans). From this simple structure, in nature, hydroxyl groups can be substituted by sugars, methyl, methoxy, and other residues, and dimers, trimers, or larger polymers are found.86 Nevertheless, in the molecular structure of resveratrol (trans3,5,4′-trihydroxystilbene) (which is the parent molecule of many stilbene oligomers, including many monomers) the 4′OH and stereoisomerism in its trans-configuration are necessary for inhibition of cellular proliferation.87 The basic units for polymerization in naturally occurring stilbene oligomers range from resveratrol to isorhapontigenin, piceatannol, oxyresveratrol, rhapontigenin, gnetol, and pterostilbene. These stilbene oligomers undergo homopolymerization and/or heteropolymerization to form a rapidly expanding body of oligostilbenes.86 Studies on the relationship between structure and efficiency in derivatives of stilbene have shown differential activity specific to the functional groups on the stilbene structure. In this study,

others (e.g., quercetin, genistein, resveratrol, or phenolic acids) cannot be considered as such. Quideau’s Definition. The current term of polyphenol has been the subject of controversy. On the basis of the recent definition provided by Stéphane Quideau (Bordeaux University, France), the term polyphenol should be used to define plant secondary metabolites derived exclusively from the shikimate-derived phenylpropanoid and/or the polyketide pathway(s) featuring more than one phenolic ring and being devoid of any nitrogen-based functional group in their most basic structural expression. A phenolic ring is a phenyl ring substituted by one or more hydroxyl groups, so the term “polyphenol” should be restricted in a strict chemical sense to structures bearing at least two phenolic moieties, irrespective of the number of hydroxyl groups they each bear. This is the reason why, for example, the shikimate-derived phenylpropanoid phenolic acid caffeic acid bearing only one phenyl ring substituted by two hydroxyl groups is not a polyphenol.16 This proposal represents a wider conception of the term polyphenol, particularly at lower molecular weights, which consequently identifies as polyphenols several thousand natural compounds. Chemical Structures and Occurrence in Nature. Polyphenols always contain heteroatom substituents besides hydroxyl groups; common substituents include ether and ester bonds as well as different carboxylic acid derivatives. In addition, ester bonds are common in the hydrolyzable tannins. More than 12000 phenolic structures have been identified, of which over 8000 are flavonoids, which all share a common backbone: two phenyl rings bound by three carbon atoms that form an oxygenated heterocycle. Flavonoids, a subclass of polyphenols with a C6 (ring A)−C3−C6 (ring B) backbone structure, based on their hydroxylation pattern and variations in the chromane heterocyclic ring (ring C), can be divided in several subclasses: chalcones, aurones, flavones, flavanols, flavanones, flavonols, anthocyanidins, and isoflavonoids. Nonflavonoids can be subdivided into: simple phenols as phenolic acids (phenylacetic acids or benzoic acids, and hydroxycinnamic acids), benzaldehyde derivatives, acetophenones, benzophenones, and polyphenols like curcuminoids, tannins (e.g., hydrolizable tannins), coumarins, xanthones, stilbenoids, lignans, and secoiridoids (see Figure 1). Polyphenols are known to associate with various carbohydrates, organic acids, and proteins.16−18 Polyphenols are found in a plethora of plants and foods.19 Although several polyphenols, such as quercetin (a flavonol), are found in many plant-based foods (cereals, legumes, fruits, tea, wine, etc.), other classes are food-type specific (e.g., flavanones in citrus, catechins in tea, isoflavones in soy, phloridzin in apples, etc.). However, different polyphenols are often found in the same food; e.g., apples contain flavanols, chlorogenic acid, glycosides of quercetin, glycosides of phloretin, and anthocyanins.20,21 The polyphenol composition can also depend on different factors such as environmental, the degree of ripeness, industrial or household processing, storage, or plant species.22−24 Relationships between Structure and Function. Recently a “periodic table” of 74 flavonoids was assembled through the use of an entropy-based algorithm (information theory), where information is defined as the negative of the logarithm of the probability distribution.77 Seven features (listed here in hierarchical order) were used to classify the flavonoids structurally: (1) the presence of ester of gallic acid and ferulic in position 3 C ring resulting in a higher stability 9417

DOI: 10.1021/acs.jmedchem.6b01026 J. Med. Chem. 2017, 60, 9413−9436

Journal of Medicinal Chemistry

Perspective

Table 1. IC50 Values for Some Selected Polyphenols in Different Growing Cancer Cellsa IC50 (μM) cancer type

cell line

Resv

Pter

EGCG

Curc

Quer

Gen

breast carcinoma

MDA-MB-231 T47D

15 ± 3 25 ± 5

63 ± 16 35 ± 11

83 ± 11 57 ± 12

12 ± 3 19 ± 4

20 ± 3 11 ± 2

43 ± 7 48 ± 12

colon carcinoma

HT29 COLO201

27 ± 4 37 ± 10

28 ± 7 29 ± 5

60 ± 9 38 ± 4

14 ± 2 18 ± 5

18 ± 3 19 ± 4

50 ± 11 73 ± 17

lung carcinoma

A549 NCI-H460

16 ± 4 12 ± 2

26 ± 4 17 ± 3

29 ± 3 18 ± 5

18 ± 4 15 ± 3

13 ± 1 12 ± 3

64 ± 9 47 ± 10

pancreas carcinoma

BxPC-3 PANC-1

14 ± 4 39 ± 7

31 ± 6 50 ± 10

33 ± 6 62 ± 13

29 ± 7 53 ± 4

16 ± 4 62 ± 11

79 ± 12 87 ± 15

melanoma

MML-1 SK-MEL-2

12 ± 3 14 ± 4

17 ± 3 17 ± 4

25 ± 7 31 ± 4

19 ± 3 36 ± 6

15 ± 3 12 ± 2

42 ± 7 36 ± 5

glioblastoma

U87 LN229

20 ± 5 11 ± 2

34 ± 7 14 ± 2

18 ± 2 15 ± 3

16 ± 5 15 ± 2

23 ± 2 12 ± 3

55 ± 8 44 ± 6

a

Cancer cells were cultured using the standard conditions recommended by the ATCC in each case.122 Cell viability was calculated using trypan blue exclusion and LDH release. Polyphenol exposure (in a scale from 0 to 100 μM) was for 48 h. IC50 values were calculated by generating a regression equation. Standard deviation of triplicate samples was calculated for three independent samples, mean ± SD (n = 3). Resv, resveratrol; Pter, pterostilbene; EGCG, epigallocatechin gallate; Curc, curcumin; Quer, quercetin; Gen, genistein.

of cancer,7,8 neuroprotection, and neurodegeneration-related pathologies,93,94 cardiovascular dysfuction and damage,95,96 metabolic syndrome,97,98 diabetes,99 aging,100,101 and inflammation-related diseases.102−104

resveratrol inhibited luminol-enhanced chemiluminescence of whole blood and activated polymorphonuclear leukocytes, whereas pinosylvin heavily affected intracellular chemiluminescence and pterostilbene extracellular chemiluminescence.88 This evidence suggest that particular functional groups in the stilbenoid structure influence (determine?) their antioxidative activity (and mechanism?). Nevertheless, recent results identify actions of polyphenols which are independent from antioxidative mechanisms. A primary example of this is epigallocatechin 3-gallate, which interacts with proteins and phospholipids in the plasma membrane and regulates signal transduction pathways, transcription factors, DNA methylation, mitochondrial function, and autophagy.89 Interestingly the abundance of polyphenols and their capacity to show a wide range of activities (estrogen, apoptotic, antidepressant, anti-inflammatory, antioxidant, pro-oxidant, etc.) make them an interesting class of chemicals to be studied using (as recently proposed) the topological substructural molecular design (TOPS-MODE) approach, which may predict not only the antioxidant activity but also the other properties discussed.90



ANTITUMOR PROPERTIES AND PROPOSED MECHANISMS At present there is abundant information regarding cellular mechanisms by which polyphenols may affect carcinogenesis,105 tumor cell proliferation and death,7 inflammation,104 angiogenesis,106 dissemination,7 and drug and radiation resistance.107 Recent excellent reviews spanning these topics abound. Basic research utilizing cancer cells have proposed several potential mechanisms for the antitumoral activity of polyphenols. Polyphenols show a wide range of antimutagenic effects. However, whether a specific polyphenol is antimutagenic or not depends on its chemical structure, the studied gene, the mutagenic factor (ultraviolet radiation, alcohol, tobacco, etc.), and whether the polyphenol is present before, during, or after exposure to the mutagen. Polyphenols exhibit antimutagenic effects maintaining genomic stability and increasing detoxification mechanisms by inducing antioxidant genes expression such as glutathione S-transferase, catalase, or quinone reductase.108 Frequently, the molecular mechanisms of polyphenols encompass induction of antioxidant enzymes, thereby reducing ROS (reactive oxygen species)-induced carcinogenesis via redox-sensitive transcription factor nuclear factor erythroid 2 p45 (NF-E2)-related factor (Nrf2), a master regulator of the transcription of many detoxifying enzymes.109,110 However, under in vivo conditions, IV administration of pterostilbene downregulates Nrf2-dependent signaling/transcription and the antioxidant protection of melanoma and pancreatic cancer cells, sensitizing the tumor to physiological prooxidant defenses and to chemotherapy.33 This biological paradox is caused, apparently, by the concentration and time differences between



LINKS WITH HUMAN HEALTH AND DISEASE Intense interest in translating the protection shown in plants to humans, in a targeted manner, has been the subject of a wide breadth of research over the last several decades. One prevailing theory proposes that natural polyphenols are at least partially responsible for the health benefits derived from consuming fruits and vegetables, but no strong evidence exists to date that dietary polyphenols actually provide health benefits.23,91,92 Understanding the effects and underlying mechanisms of polyphenols within the human organism, and their clinical outcome in human diseases, remains a matter of controversy in nutrition science and disease prevention. Despite this, and based on strong experimental support, polyphenols may be beneficial (but not limited to) in the prevention and treatment 9418

DOI: 10.1021/acs.jmedchem.6b01026 J. Med. Chem. 2017, 60, 9413−9436

Journal of Medicinal Chemistry

Perspective

Table 2. Effect of Some Selected Polyphenols, at Bioavailable Concentrations, on Different Cancer Cells Growing in Vitroa % of viable cells relative to controls cell line

Resv (16 μM)

Pter (34 μM)

EGCG (6 μM)

Curc (41 μM)

Quer (14 μM)

Gen (18 μM)

breast carcinoma

MDA-MB-231 T47D

95 ± 3 99 ± 5

90 ± 3 91 ± 7

102 ± 6 100 ± 7

90 ± 7 88 ± 7*

103 ± 4 100 ± 7

91 ± 4 93 ± 3

colon carcinoma

HT29 COLO201

98 ± 4 96 ± 5

96 ± 6 95 ± 5

96 ± 9 100 ± 6

83 ± 6** 81 ± 8**

94 ± 5 101 ± 5

104 ± 7 100 ± 9

lung carcinoma

A549 NCI-H460

93 ± 5 94 ± 7

81 ± 9** 83 ± 8*

103 ± 4 105 ± 5

96 ± 3 95 ± 5

104 ± 2 95 ± 4

107 ± 6 102 ± 8

pancreas carcinoma

BxPC-3 PANC-1

97 ± 4 95 ± 7

102 ± 4 99 ± 5

96 ± 7 102 ± 8

96 ± 5 99 ± 3

melanoma

MML-1 SK-MEL-2

102 ± 6 99 ± 3

105 ± 7 97 ± 6

103 ± 7 97 ± 6

104 ± 7 105 ± 4

glioblastoma

U87 LN229

98 ± 5 106 ± 4

92 ± 7 91 ± 7

97 ± 5 106 ± 5

98 ± 5 100 ± 5

cancer type

98 ± 5 105 ± 4 94 ± 6 95 ± 7 99 ± 8 102 ± 4

97 ± 6 94 ± 3 87 ± 4** 85 ± 7* 103 ± 5 98 ± 5

a Independent experiments were run for each polyphenol (n = 5 IV, and n = 5 oral). For the IV administration, each polyphenol was dissolved in DMSO:ethanol (2:1), and the total volume administered was never >30 μL. For the oral administration, each polyphenol was dissolved in a selfmicroemulsifying drug delivery system comprised of 60% surfactant (emulsifier OP:Cremorphor EL = 1:1), 30% cosurfactant (PEG 400), and 10% oil (ethyl oleate),123 whereas the total volume administered was never >100 μL. Plasma levels of Resv, Pter, and Quer were measured as previously described.68,124 Plasma levels of EGCG were measured as described by Lambert et al.125 Plasma levels of Curc were measured as described by Zhongfa et al.126 Plasma levels of Gen were measured as described by Ju et al.127 After IV administration, peak level of (a) Resv was of 32 ± 4 μM at 5 min, (b) Pter was of 67 ± 7 μM at 5 min, (c) EGCG was of 12 ± 3 μM at 5 min, (d) Curc was of 81 ± 9 μM at 5 min, (e) Quer was of 27 ± 5 μM at 5 min, and (f) Gen was of 35 ± 4 μM at 5 min. Plasma levels of all these PFs were