Flavonoids and the lifespan of the worm Caenorhabditis elegans

This nematode is widely used as a model organism since it is easy to culture under laboratory conditions and very well characterised. Caenorhabditis elegans was the first multicellular organism to have its genome fully sequenced⁽ Reference Consortium ^{26 )} and many of its gene functions have been characterised by generating mutant strains. Approximately 60–80 % of human genes have homologues in C. elegans, which, similar to human subjects, is affected by ageing. In C. elegans, hundreds of genes that modulate longevity have been identified⁽ Reference Tissenbaum ^{27 )}. Information on the worm genome and gene functions can be found on the website www.wormbase.org. C. elegans is a self-fertilising hermaphrodite, which develops rapidly passing through four larval stages with a generation time of 3 d (www.wormbook.org). Because of its multicellularity and its different tissues and cell types such as neurons, gut and muscle cells, the nematode is a suitable model organism to study metabolic functions and other processes related to ageing such as neurodegeneration. C. elegans can be maintained from about 12 to 25°C on agar with Escherichia coli as its diet (www.wormbook.org). At 25°C the worm has a life expectancy of approximately 15 d, with a maximum lifespan of approximately 22 d⁽ Reference Friedman and Johnson ^{28 )}. Such a short lifespan makes it a useful organism for longevity studies.

In our search, we found a total of twenty-seven reports on flavonoids or flavonoid-containing plant extracts and their impact on C. elegans lifespan. Three of these studies dealt with catechin, two studies were on epicatechin, one study on methylated epicatechins, three studies on EGCG, nine studies on quercetin, three studies on glycosylated quercetins, two on methoxylated quercetins, two on kaempferol, one on kaempferol derivatives, two on myrecitin, one on naringenin, one on isoxanthohumol, one on baicalein, one on flavolignan and seven on plant extracts (Table 1).

Table 1.

Lifespan and flavonoids in Caenorhabditis elegans

* If the numbers in the control and treatment groups are different, the lower number is given. If various trials were carried out the sum of the animals used is given; L-larval.

The flavanol catechin increases C. elegans lifespan in some settings. Saul et al.  found that catechin extended lifespan of the wild-type N2 strain at concentrations from 100 to 800 µm and temperatures ranging from 15 to 23°C⁽ Reference Saul, Pietsch and Menzel ^{29 ,} Reference Saul, Pietsch and Sturzenbaum ^{30 )}. However, Surco-Laos et al. ⁽ Reference Surco-Laos, Duenas and Gonzalez-Manzano ^{31 )} could not reproduce this finding when they tested the effect of 200 µm quercetin at 20°C in the same worm strain. In contrast to Saul et al. who used the F₁ generation of parents that had been exposed to catechin, Surco-Laos et al. started the exposition of the worm to the flavanol later in life (larval stage 1).

The findings by Saul et al. were followed by studies in mutant worms. Interestingly, catechin did not extend lifespan in the mutant strains daf-2 (the orthologue of the human insulin/insulin-like growth factor (IGF) receptor), nhr-8 (a nuclear hormone receptor), akt-2 (homologue of the serine/threonine kinase Akt/protein kinase B), mev-1 (a subunit of the mitochondrial respiratory chain complex II, the worms have increased levels of oxidative stress) and eat-2 (worms carrying this mutation show decreased pharyngeal feeding and therefore are energy restricted).

It remains unclear whether epicatechin benefits longevity of worms. While Bartholome et al. ⁽ Reference Bartholome, Kampkotter and Tanner ^{32 )} reported lifespan extension at 100 µm epicatechin in the feed, Surco-Laos et al. did not observe any changes using the concentration of 200 µm. However, they found that the two derivatives of epicatechin, which are methylated at one of the hydroxyl groups of the B ring (3′-O-methylepicatechin and 4′-O-methylepicatechin) increased C. elegans lifespan.

EGCG, the ester of gallocatechin (a flavanol with three hydroxyl groups in the B ring) with gallic acid also appears to prolong lifespan if supplemented at concentrations from 100 to 220 µm ⁽ Reference Bartholome, Kampkotter and Tanner ^{32 ,} Reference Abbas and Wink ^{33 )}. At concentrations of 22 µm and lower there were no effects observed under normal growth conditions⁽ Reference Zhang, Jie and Zhang ^{34 )}. Interestingly, EGCG could also prolong lifespan in mev-1 mutants.

Of the nine studies on quercetin, seven reported quercetin as prolonging lifespan at concentrations ranging from approximately 70 to 200 µm ⁽ Reference Kampkotter, Timpel and Zurawski ^{35 –} Reference Duenas, Surco-Laos and Gonzalez-Manzano ^{41 )}. One study supplementing quercetin at 166 µm could not detect lifespan extension⁽ Reference Wu, Smith and Paramasivam ^{42 )}, another study applying quercetin at 50 and 250 µm reported no change and a decrease in lifespan, respectively⁽ Reference Pietsch, Saul and Chakrabarti ^{43 )}. Furthermore, quercetin supplementation was tested in various mutant strains. These studies concluded that mutations in sek-1 (a mitogen-activated protein kinase kinase), daf-2, age-1 (the orthologue of the human phosphoinositide-3-kinase p110 catalytic subunit) and unc-43 (the orthologue of the human type II calcium/calmodulin-dependent protein kinase) abrogated lifespan extension⁽ Reference Pietsch, Saul and Chakrabarti ^{43 )}. Pietsch et al. also found that quercetin could prolong the lifespan of mev-1 mutants, which experience high levels of oxidant stress. Contrarily, Grunz et al. ⁽ Reference Grunz, Haas and Soukup ^{40 )} found no lifespan increase in quercetin supplemented mev-1 mutants.

Beside its concentration, the glycosylation of quercetin also seems to influence lifespan extension. At position 3 or 3′ glycosylated quercetin prolonged lifespan when used at concentrations of 10 and 25 µm (quercetin-3-O-glycoside)⁽ Reference Duenas, Surco-Laos and Gonzalez-Manzano ^{41 )} or 66 µm (quercetin-3′-O-glycoside)⁽ Reference Xue, Ahiko and Miyakawa ^{39 )}. In contrast, higher concentrations of quercetin-3-O-lycoside (50 and 100 µm) had no effect and 200 µm even shortened lifespan⁽ Reference Duenas, Surco-Laos and Gonzalez-Manzano ^{41 )}. The authors who also observed that quercetin at 100 µm increased lifespan explain this finding with a higher uptake and bioavailability of the glycosylated form. In the worm, the glycoside is metabolised into the active quercetin by deglycosylation. The klo-1 and klo-2 mutant worms with defects in β-glucosidase activity benefited from quercetin-3-O-glucoside concentrations as high as 200 µm possibly the lack of β-glucosidase prevented the formation of quercetin from quercetin-3-O-glycoside at detrimental levels. Furthermore, the type of sugar that is conjugated to quercetin also seems to be important for lifespan increase. While quercetin that was glycosylated at position 3 with a disaccharide consisting of two glucose monomers (Q3M) or two rhamnose monomers (quercetin-3-O-α-L-rhamnopyranosyl(1→2) O-α-L-rhamnopyranoside) increased lifespan in N2 worms, rutin, which is glycosylated at the same position with the disaccharide rutinose, could not prolong longevity⁽ Reference Xue, Ahiko and Miyakawa ^{39 ,} Reference Ahn, Lee and Kim ^{44 )}. Similarly to the aglycon, quercetin with a single methylation of either hydroxyl group in the B ring (isorhamnetin: 3′methylation, tamarixetin: 4′ methylation) extends lifespan in worms at 200 µm ⁽ Reference Surco-Laos, Cabello and Gomez-Orte ^{38 )} and tamarixetin also at 158 µm ⁽ Reference Wu, Smith and Paramasivam ^{42 )}.

For kaempferol we found two studies that reported contradicting data. While the study that used the higher concentration of 175 µm did not observe lifespan extension in supplemented N2 worms⁽ Reference Wu, Smith and Paramasivam ^{42 )}, the second study using 100 µm kaempferol observed a slight lifespan increase, which was absent in the mev-1 mutant⁽ Reference Grunz, Haas and Soukup ^{40 )}. The kaempferol derivative icariin prolonged lifespan at 45 µm but not at 15 or 75 µm and not in daf-2 or daf-16 (forkhead box O (FOXO) homologue) mutants. While deglycosylation of icariin, which is glycosylated at positions 3 and 7, at both (icaritin) or at the 3 position (icariside I) seems to cancel out its lifespan-expanding properties, the derivative that is deglycosylated at position seven (icariside II) extended lifespan at 20 µm but not at 10 or 40 µm. In studies with daf-2, daf-16, hsf-1, rsks-1 and eat-2 mutants, the impact of icariside II depended on the insulin/IGF receptor orthologue (daf-2), the FOXO homologue (daf-16), the heat-shock transcription factor orthologue (hsf-1) but not on a putative ribosomal protein S6 kinase (rsks-1). This was also observed in energy restricted worms (eat-2)⁽ Reference Cai, Huang and Zhang ^{45 )}.

Myricetin was used in two studies and both the studies concluded that it prolonged lifespan at 100 μm. This seems to depend on the C. elegans FOXO orthologue⁽ Reference Buchter, Ackermann and Havermann ^{46 )}, whereas mev-1 mutants also benefitted from myricetin supplementation⁽ Reference Grunz, Haas and Soukup ^{40 )}.

The flavanones naringenin and isoxanthohumol did not promote longevity in the worm at 100 or 200 µm and in FOXO mutants, 200 µm isoxanthohumol even shortened lifespan⁽ Reference Grunz, Haas and Soukup ^{40 ,} Reference Buchter, Havermann and Koch ^{47 )}.

In contrast, the flavone baicalein prolonged lifespan at 100 µm in wild-type worms⁽ Reference Havermann, Rohrig and Chovolou ^{48 )}. In the case of the flavolignan isomers silymarin, the effect appeared to be dose dependent, since 25 and 50 µm increased the worm's lifespan, while 100 µm shortened it⁽ Reference Kumar, Park and Awasthi ^{49 )}.

All seven studies on plant extracts reported positive outcomes on wild-type worm lifespan. The extracts studied were a polyphenol-enriched cocoa powder containing approximately 120 g/kg procyanidins⁽ Reference Martorell, Forment and de Llanos ^{50 )}, a hot water extract from Taiwan hinoki containing catechin, quercetin, quercetin-3-O-rhamnoside and myricetin-3-O-rhamnoside⁽ Reference Cheng, Li and Shi ^{51 )}, a tea seed pomace extract containing approximately 44 % kaempferol glycosides⁽ Reference Wei, Yu and Yen ^{52 )}, the total flavones from nymphea hybrid⁽ Reference Zhuang, Lv and Li ^{53 )}, the ginkgo biloba extract EGb761 (24 % flavone glycosides)⁽ Reference Wu, Smith and Paramasivam ^{42 )}, a proanthocyanidin-enriched fraction from a blueberry extract (BE)⁽ Reference Wilson, Shukitt-Hale and Kalt ^{54 )} and an anthocyanin-rich extract from purple wheat-containing cyanidin-3-O-glycoside, peonidin3-O-glucoside and malvidin-3-O-galactoside⁽ Reference Chen, Muller and Richling ^{55 )}. The cocoa powder and the purple wheat extract were also tested in mutants where the effect of the procyanidin-rich cocoa powder depended on daf-16, sir2.1 and daf-2 ⁽ Reference Martorell, Forment and de Llanos ^{50 )}. Similarly, the anthocyanin-rich wheat extract did not prolong lifespan in daf-16 mutants. However, this extract extended lifespan in the more sensitive towards oxidative stress than wild-type worms, the mev-1 mutants⁽ Reference Chen, Muller and Richling ^{55 )}, which were not tested with the cocoa powder. Interestingly, some of the extracts such as the tea seed pomace⁽ Reference Wei, Yu and Yen ^{52 )} or the Taiwan hinoki extract⁽ Reference Cheng, Li and Shi ^{51 )} extended lifespan at concentrations that were as low as 1–2 mg/l.

Most studies in C. elegans report positive effects on worm lifespan when used at μm concentrations, however, with the concentration depending on the flavonoid supplemented, high doses may shorten lifespan. Interestingly, not all flavonoids seem to affect the same molecular targets. While the action of kaempferol, its derivatives and myricetin depended on a functional FOXO homologue, quercetin and catechin increased lifespan irrespectively of daf-16. However, IGF receptor mutants (daf-2) did not live longer when supplemented with any of the flavonoids tested.

Flavonoids and the lifespan of Drosophila melanogaster

Similar to C. elegans, the fruit fly Drosophila melanogaster is a multicellular organism that is easy to house and extensively used as a model organism. Its organs, development and behaviour show similarities to human subjects. The fly has a relatively short lifespan of 2–3 months and shows symptoms of ageing⁽ Reference He and Jasper ^{56 )}. Its genome has been fully sequenced⁽ Reference Adams, Celniker and Holt ^{57 )} and information on genes and mutants has been deposited on www.flybase.org.

We found ten studies on flavonoids and lifespan studies in D. melanogaster. Three of them used single compounds such as the flavanol epicatechin and the flavonol fisetin⁽ Reference Wood, Rogina and Lavu ^{58 –} Reference Wagner, Piegholdt and Rabe ^{60 )}, the other seven studies were on cocoa powder, green tea extract, Ludwigia octovalvis extract, apple polyphenols, black tea extract (BTE) and BE ⁽ Reference Li, Chan and Huang ^{61 –} Reference Lin, Chen and Wang ^{67 )} (Table 2).

Table 2.

Lifespan and flavonoids in Drosophila melanogaster

EGCG, epigallocatechin gallate; CSY, cornmeal–sugar–yeast.

* If the numbers in the control and treatment groups are different, the lower number is given. If various trials were carried out the sum of the animals used is given.

Epicatechin at 100–8000 µm increased fly lifespan, while 10 µm had no effect⁽ Reference Si, Fu and Babu ^{59 )}. The other single compound tested in flies, the flavonol fisetin, was tested as part of a standard diet containing 3 % yeast and an energy restricted diet with 2 % yeast. Interestingly, 10 µm fisetin did not change the lifespan of male or female flies on a 3 % yeast diet, but 100 µm fisitin in the 3 %yeast diet increased lifespan in both sexes. However, on a 2 % yeast diet female flies did not benefit from fisetin at 10 or 100 µm and male flies lived for a shorter period when supplemented with fisetin as compared with the 2 % yeast diet control. The authors hypothesise that fisitin may mimic the ER effect on lifespan and thus not be effective when used as part of an energy restricted diet⁽ Reference Wood, Rogina and Lavu ^{58 )}. EGCG increased the lifespan of male flies at 10 g/l diet. However, the authors also state that EGCG did not affect lifespan in female flies⁽ Reference Wagner, Piegholdt and Rabe ^{60 )}.

Cocoa powder improved fly longevity at 50 g/l but showed no effect at 100 g/l⁽ Reference Bahadorani and Hilliker ^{62 )}. Although the flavonoid content of this powder had not been analysed as part of the study, we included it because cocoa powder is known to have high flavonoid content⁽ Reference Lee, Kim and Lee ^{68 )}. Bahdorani and Hilliker also used mutant strains for the superoxide dismutases (SOD) 1 and 2 and observed that in animals deficient in SOD1, which is also referred to as Cu/ZnSOD and found in the cytosol of the cell, both concentrations of cocoa powder (50 and 100 g/l) increased lifespan, while in animals deficient in SOD2, also referred to as MgSOD and found in the mitochondrial matrix, lifespan was shortened.

Green tea extract (GTE) containing 62 % EGCG, 19 % epigallocatechin, 9 % epicatechin gallate, 7 % epicatechin extended lifespan in male Oregon-R-C on a diet with 10·5 % cornmeal, 2·1 % yeast and 10·5 % dextrose (m/V). The extract was tested at 1; 5 and 10 g/l with 5 and 10 g/l extending lifespan significantly⁽ Reference Li, Chan and Huang ^{61 )}. Adding fat to fly diets shortens lifespan dose-dependently⁽ Reference Li, Chan and Yao ^{69 )}. However, adding the 10 g/l GTE to a 5 or 10 % lard-containing diet improved longevity compared with the non-supplemented high-fat diet⁽ Reference Li, Chan and Yao ^{69 )}.

When using a BTE containing 50 % theaflavins, 12 % epicatechin, 12 % epigallocatechin, 10 % EGCG and 6 % epicatechin gallate with a similar diet (glucose instead of dextrose) and the same type of flies at 5 and 10 g/l, lifespan was also increased. Furthermore, BTE at 10 g/l could counteract, in part, the detrimental effect of 10 % lard addition to the diet on fly lifespan⁽ Reference Peng, Chan and Li ^{63 )}.

Ludwigia octovalvis is used as herbal medicine and consumed in parts of Asia as a drink⁽ Reference Lin, Chen and Wang ^{67 )}. Its extract containing 90 mg flavonoids/g (measured as rutin equivalents) prolonged the lifespan of male and female Canton-S or w¹¹¹⁸ flies at concentrations from 0·05 – 1 g/l when applied with a high-energy diet (15 % dextrose and 15 % yeast). Supplementation of the low-energy diet (5 % dextrose and 5 % yeast) at 0·1 g/l did not affect lifespan in w¹¹¹⁸ flies⁽ Reference Lin, Chen and Wang ^{67 )}.

Apple polyphenols containing 4 % proanthocyanidins, 3 % epicatechin, 1 % catechin and other unidentified proanthocyanidins and phenolics increased the lifespan of male Oregon-R-C flies fed on 10·5 % cornmeal, 2·1 % yeast and 10·5 % glucose (m/V) when supplemented at 10 g/l, whereas 2 g/l did not improve longevity⁽ Reference Peng, Chan and Huang ^{64 )}.

BE containing 49 % cyanidin-3-O-glucoside and 20 % petunidin-3-O-glucoside and used with the same type of flies and diet as the apple polyphenols increased the lifespan supplemented at a concentration of 5 g/l, while 2 g/l did not change lifespan⁽ Reference Peng, Zuo and Kwan ^{65 )}.

Black rice extract containing lower levels of cyanidin and petunidin glycosides than the BE (24 % cyanidine-3-O-glucoside and 1·5 % petunidin-3-O-glucoside) extended lifespan in the same type of flies as the BE and apple polyphenols at a sixfold higher concentration (30 g/l) than BE, which is also high in anthocyanins. The twofold higher concentration (10 g/l) did not affect lifespan⁽ Reference Zuo, Peng and Liang ^{66 )}. This finding is consistent with the BE having no effect at 2 g/l since the black rice extract has less than half of the cyanidin content and an approximately sevenfold lower petunidin content compared with the BE.

Overall, experiments in flies showed a beneficial effect of flavonoids or flavonoid-containing extracts on the lifespan of D. melanogaster even with concentrations in the feed that were sixtyfold higher than in the worm. Of interest, the diet composition strongly influenced the outcome of the studies, since energy restricted diets abrogated or reversed the flavonoid effect on longevity while high-energy diets increased it.

Flavonoids and the lifespan of Mus musculus

Being mammals and vertebrates, mice share even more similarities with human subjects than worms and flies. Human homologues have been found for approximately 98 % of mouse genes⁽ Reference Mural, Adams and Myers ^{70 )} and many genetic functions have been studied in knockout mice. Compared with other mammals, mice are easy to house and have a short lifespan of 2–3 years. This makes them a useful model organism for ageing research⁽ Reference Selman and Withers ^{71 )}.

Of six mouse trials that studied the influence of flavonoid-containing extracts on lifespan, two studies also tested the flavonols quercetin and icariin as single compounds (Table 3).

Table 3.

Lifespan and flavonoids in Mus musculus

BE, blueberry extract; GTE, green tea extract; PG, pomegranate powder.

* If the numbers in the control and treatment groups are different, the lower number is given. If various trials were carried out the sum of the animals used is given.

While icariin extended mean lifespan at 0·02 % diet when fed to C57BL/6 from age 12 months but did not extend maximum lifespan⁽ Reference Zhang, Cai and Huang ^{72 )}, feeding of quercetin at a fivefold higher dose than icariin (0·1 % diet) and beginning at age 5 weeks decreased the lifespan of LACA strain mice⁽ Reference Jones and Hughes ^{73 )}. Two out of six studies testing flavonoid-containing extracts observed a lifespan increase in supplemented mice. These extracts were a triple combination of BE, GTE and pomegranate powder (PG) in the study by Aires et al. ⁽ Reference Aires, Rockwell and Wang ^{74 )} and green tea polyphenols containing approximately 70 % epigallocatechins, epicatechins and gallocatechins in a study by Kitani et al. ⁽ Reference Kitani, Osawa and Yokozawa ^{75 )}. Interestingly, the combination of BE, EGCG and PG was compared with a non-supplemented high-fat diet in C57BL/6 mice that had access to feed for 24 h followed by 24 h fasting (intermittent fasting). While intermittent fasting compared with ad libitum feeding increased lifespan when started at age 20 weeks, a 2 % BE, 0·015 % GTE and 0·3 % PG diet further prolonged lifespan in intermittent fasting-fed mice⁽ Reference Aires, Rockwell and Wang ^{74 )}. The green tea polyphenols were tested in 13-month-old C57BL/6 on an ad libitum moderate fat diet. By supplementing their drinking water at 80 mg/l the mice lived longer than the unsupplemented controls⁽ Reference Kitani, Osawa and Yokozawa ^{75 )}. However, four studies testing blackcurrant juice (containing anthocyanins, quercetin and quercetin glycosides)⁽ Reference Jones and Hughes ^{73 )}, epimedium flavonoids (containing 20 % icariin)⁽ Reference Zhang, Cai and Huang ^{72 )}, GTE⁽ Reference Strong, Miller and Astle ^{76 )}, BE, GTE, a triple combination of GTE, BTE and morin, PG or a triple combination of quercetin, taxifolin and pynogenol^{® (} Reference Spindler, Mote and Flegal ^{77 )} did not detect lifespan increase in the mice.

In mice, it remains unclear whether flavonoid supplementation could prolong lifespan. While in C57BL/6 mice there are indications that these mice could benefit from elevated flavonoid intake, genetically heterogenous mice do not live longer when supplemented. Moreover, a very high flavonoid dose (1 g/kg diet) shortened the lifespan of LACA mice.

Factors putatively affecting lifespan modulation by flavonoids

Although flavonoids that extend worm lifespan do not necessarily expand mammalian lifespan, experiments in worms and flies can help to understand which proteins or signalling pathways are affected by the flavonoids. Thus, invertebrate models could be a useful tool to investigate structure–activity relationships. As can be deduced from studies in mutant worms and flies (Tables 1 and 2), flavonoids affect various molecular targets and may influence lifespan by numerous mechanisms (Fig. 2). Flavonoid pleiotropy and human subjects consuming many different flavonoids with their diet and not just single compounds complicate the evaluation of flavonoids in the context of lifespan extension.

Fig. 2.

(Colour online) Putative pleiotropic effects by which flavonoids may affect lifespan in model organisms.

Extracts or single compounds

It remains unclear whether different flavonoids that are consumed at the same time have additive or even synergistic effects on lifespan. It is also possible that other compounds in the diet attenuate the effect of life-prolonging flavonoids. To test this, studies experimenting with single compounds and combinations of flavonoids and/or other secondary plant metabolites are needed. Although there are studies that tested a flavonoid-containing extract and single flavonoids from this extract, the concentrations in the extracts are different from the tested dose of the single compounds. This makes it difficult to deduce whether varying results are due to changes in concentration or the combination with other polyphenols. For a ginkgo biloba extract, the authors found that one of its flavonoids, kaempferol, did not affect lifespan on its own while isolated tamarixetin prolonged lifespan to a greater extent than the extract. This could have occurred because the tamarixetin dose the worms were exposed to was higher in the single compound experiment than in the ginkgo extract experiment. Alternatively, components of the extract could have attenuated the lifespan-prolonging action of tamarixetin or may be slightly toxic⁽ Reference Wu, Smith and Paramasivam ^{42 )}.

Interestingly, some of the extracts used in worms such as the tea seed pomace or the Taiwan hinoki extract increased lifespan at lower concentrations than single compounds⁽ Reference Cheng, Li and Shi ^{51 ,} Reference Wei, Yu and Yen ^{52 )}. Assuming these extracts consisted entirely of a flavonoid with the molecular mass of catechin, 1–2 mg/l would correspond to 3·5–7 µm flavonoid/l. If extracts worked at lower doses than single compounds this might be due to synergisms between different flavonoids and/or other polyphenols. In flies, the concentrations used to test the extracts tended to be higher than the single compounds and manifold higher than in the worms. In wild-type flies approximately 10 g flavonoids/l seemed non-toxic⁽ Reference Li, Chan and Huang ^{61 ,} Reference Peng, Chan and Li ^{63 )}. However in flies, only epicatechin and fisetin were tested on their own, making it difficult to hypothesise whether flavonoids in extracts show additive, synergistic or antagonistic actions.

In mice, blackcurrant juice was tested and did not alter lifespan, but when testing one of the flavonols in blackcurrant juice, quercetin, a manifold higher concentration shortened the lifespan of the mice⁽ Reference Jones and Hughes ^{73 )}. When testing epimedium flavonoids of which 20 % are icariin at 0·06 %, these did not prolong mouse lifespan whereas, when supplementing icariin at 0·02 %, the mice lived longer⁽ Reference Zhang, Cai and Huang ^{72 )}. Comparing icariin intake of a 30 g mouse consuming 5 g feed daily to an estimated human intake of 70 mg flavonoids daily⁽ Reference Beecher ^{1 )}, the intake/kg bodyweight is  thirty times higher in the mouse. Thus, the lifespan extending concentrations of flavonoids are not consumed in a non-supplemented human diet. Moreover, the toxic dose (0·1 %) of quercetin in mice⁽ Reference Jones and Hughes ^{73 )} is over 100 times higher than the estimated human intake in mg/kg bodyweight.

Antimicrobial properties

Antimicrobial, antioxidative and ER-like mechanisms have been proposed to prolong lifespan in flavonoid-supplemented animals. It has been shown that flavonoids and other polyphenols possess antimicrobial properties and that feeding worms dead bacteria as opposed to live bacteria prolongs lifespan. When supplementing gallic acid or ellagic acid to heat-killed bacteria no further lifespan enhancing effect is observed because of the polyphenols although both extend lifespan when fed as part of a live bacteria diet⁽ Reference Saul, Pietsch and Sturzenbaum ^{30 ,} Reference Kim, Kim and Choi ^{78 )}. Interestingly, catechin and quercetin extended lifespan also when the worms were fed dead bacteria⁽ Reference Saul, Pietsch and Menzel ^{29 ,} Reference Pietsch, Saul and Menzel ^{37 )}, thereby making it unlikely that these flavonoids promote longevity because of their antimicrobial properties. Furthermore, it was shown in mice and rats that flavonoids influenced the gut microbiota, which could benefit the host and as a consequence promote longevity⁽ Reference Massot-Cladera, Abril-Gil and Torres ^{79 ,} Reference Anhe, Roy and Pilon ^{80 )}.

Antioxidant properties

In worms, it has been shown that flavonoids decrease oxidative stress⁽ Reference Wei, Yu and Yen ^{52 ,} Reference Fitzenberger, Deusing and Marx ^{81 )}. Although flavonoids can work as radical scavengers because of their chemical structure, their in vivo antioxidative action is more likely caused by them stabilising and up-regulating the antioxidant transcription factor nuclear factor (erythroid-derived 2)-like 2. Nuclear factor (erythroid-derived 2)-like 2  activation leads to the expression of its target genes that remove peroxides and eliminate xenobiotics⁽ Reference Tanigawa, Fujii and Hou ^{17 )}. Some studies found that EGCG could increase the lifespan of oxidatively stressed worms⁽ Reference Zhang, Jie and Zhang ^{34 )}. Additionally, experiments in mev-1 mutants that experience high levels of oxidative damage and have a shorter lifespan than wild-type worms, showed that myricetin, quercetin and purple wheat extract could prolong their lifespan⁽ Reference Saul, Pietsch and Menzel ^{36 ,} Reference Buchter, Ackermann and Havermann ^{46 ,} Reference Chen, Muller and Richling ^{55 )}. However, quercetin does not always benefit mev-1 worms and kaempferol did not increase lifespan in the mutants⁽ Reference Grunz, Haas and Soukup ^{40 )}. In flies, cocoa powder was also tested in SOD-deficient mutants. Interestingly, it prolonged lifespan in flies that did not have the cytosolic SOD1, but was detrimental for flies without the mitochondrial SOD2. These findings point to the notion that when oxidative stress levels are very high, flavonoids could further increase oxidative damage⁽ Reference Bahadorani and Hilliker ^{62 )}. Moreover, it has been hypothesised that low levels of oxidative stress may increase the lifespan of model organisms by activating their antioxidant defence⁽ Reference Schulz, Zarse and Voigt ^{82 )}. Thus, flavonoids may not only activate nuclear factor (erythroid-derived 2)-like 2  by reacting with its inhibitor Kelch-like ECH-associated protein 1 and thereby allowing the transcription factor to locate to the nucleus⁽ Reference Dinkova-Kostova, Holtzclaw and Cole ^{83 )}, but also by producing low levels of oxidative stress, which in turn induce the transcription of antioxidative enzymes. The observation that a low dose of a stimulus has a beneficial and a high dose of a stimulus has the opposite (negative) effect has been described as hormesis⁽ Reference Mattson ^{84 )}. In experiments with worms, it was shown that lifespan extension by quercetin was dose dependent with low doses prolonging lifespan and higher doses shortening lifespan⁽ Reference Duenas, Surco-Laos and Gonzalez-Manzano ^{41 ,} Reference Pietsch, Saul and Chakrabarti ^{43 )}. Interestingly, this dose–response curve may not apply to all flavonoids since catechin prolonged worm lifespan to a very similar extent over a concentration range of 100–800 µm ⁽ Reference Saul, Pietsch and Sturzenbaum ^{30 )}. The finding that some flavonoids are toxic at high doses and others do not appear to harm the organism even at higher concentrations points to the notion that different flavonoids affect lifespan via different mechanisms⁽ Reference Pallauf and Rimbach ^{85 )}. In the studies reviewed on quercetin and worms (Table 1), lifespan extension depended on methyl groups or sugars conjugated to the flavonoid⁽ Reference Surco-Laos, Cabello and Gomez-Orte ^{38 ,} Reference Xue, Ahiko and Miyakawa ^{39 ,} Reference Wu, Smith and Paramasivam ^{42 )}. Apart from changing the interaction with molecular targets inside the organism these modifications also affect the bioavailability of the flavonoid⁽ Reference Duenas, Surco-Laos and Gonzalez-Manzano ^{41 )}.

Studies in mutants

Studies in other mutant worms also showed that the mechanisms underlying the lifespan increase might differ depending on the flavonoid. The longevity-promoting impact of icariin, icariside II, quercetin, catechin and cocoa powder was not triggered in daf-2 mutant worms (Table 2). None of the flavonoids or extracts tested prolonged the lifespan of this mutant lacking the human IGF homologue. However, in an age-1 mutant (phosphoinositide-3-kinase subunit p110 homologue) catechin increased lifespan, while quercetin did not. In turn, catechin failed to increase lifespan in an akt-2 (AKT homologue) mutant, while quercetin could prolong lifespan in these worms⁽ Reference Saul, Pietsch and Sturzenbaum ^{30 ,} Reference Pietsch, Saul and Menzel ^{37 )}. In contrast to quercetin and catechin, the life-expanding impact of icariside II, myricetin, cocoa powder and purple wheat extract was abolished by mutating daf-16, the worm FOXO homologue⁽ Reference Cai, Huang and Zhang ^{45 ,} Reference Buchter, Ackermann and Havermann ^{46 )}. Furthermore, the effect of cocoa powder but not quercetin or catechin depended on sir-2.1, a histone deacetylase which is similar to mammalian SIRT1⁽ Reference Martorell, Forment and de Llanos ^{50 )}.

Energy-restriction-like effects

Oxidative stress and IGF signalling are reduced in animals undergoing ER. It could be hypothesised that flavonoids mimic, in part, the effects or ER, which is known to prolong lifespan in worms, flies and mice⁽ Reference Nikolai, Pallauf and Huebbe ^{14 )}. However, while catechin failed to prolong lifespan in a worm that was GM to reduce its feed intake, quercetin and icariside II prolonged lifespan in an ER worm model⁽ Reference Saul, Pietsch and Menzel ^{29 ,} Reference Pietsch, Saul and Menzel ^{37 ,} Reference Cai, Huang and Zhang ^{45 )}. Interestingly, in mice, intermittent fasting and supplementing a combination of GTE, BE and PG had additive effects on lifespan increase. In contrast in flies, fisetin did not extend the lifespan of flies fed an ER diet, but as part of a diet with higher yeast content it could prolong lifespan. Flavonoids could protect an organism from the negative impact of overeating since BTE and GTE partly attenuated the lifespan decrease caused by feeding flies fat-containing diets. In worms, GTE was also shown to prevent glucose-induced survival reduction⁽ Reference Deusing, Winter and Kler ^{86 )}. In a mouse study by Spindler et al.⁽ Reference Spindler, Mote and Flegal ^{77 )}, although ER extended lifespan, the tested flavonoid-containing extracts did not improve longevity. Of interest, the control group was fed isoenergically compared with the group receiving the flavonoid extracts because the researchers observed a lower feed intake in the supplemented group. Therefore, in future studies feed intake should be monitored carefully in order to detect whether the flavonoid or a flavonoid-caused decrease in food intake extends lifespan of the model organism. However, since antinutritional qualities of flavonoids have been described, it is possible that flavonoids mimic ER by binding to proteins and inhibiting enzymes⁽ Reference Rawel, Meidtner and Kroll ^{87 ,} Reference Wiese, Gartner and Rawel ^{88 )}, e.g. lipases⁽ Reference Buchholz and Melzig ^{89 )}, α-amylase, α-glucosidase⁽ Reference Wagner, Piegholdt and Rabe ^{60 ,} Reference Ong and Khoo ^{90 ,} Reference Wang, Du and Song ^{91 )}, proteases such as trypsin⁽ Reference Shahwar, Raza and Atta Ur ^{92 )} or glucose transporters⁽ Reference Noteborn, Jansen and Benito ^{93 ,} Reference Ader, Block and Pietzsch ^{94 )}. Moreover, flavonoids may affect the microbiota composition⁽ Reference Massot-Cladera, Abril-Gil and Torres ^{79 )} and thus exert ER-like effects via interaction with the microbiota⁽ Reference Esposito, Damsud and Wilson ^{95 )}.

Mice do not always benefit from flavonoid supplementation

The negative outcomes of most mouse studies show that the lifespan-extending effects of flavonoids in worms and flies need to be verified in mammals since mice seem less likely to respond to flavonoid supplements. Moreover, flavonoid metabolism may differ between species.

In contrast to the findings in worms, quercetin supplemented as a single compound decreased mouse lifespan and diets with approximately 40 or 580 mg quercetin/kg diet did not change lifespan. However, in the case of icariin, studies in mice and worms showed positive effects. The data on GTE and mouse longevity are contradicting. In one out of three studies GTE extended lifespan, whereas in the other two no differences were observed. In addition to differences in the dose, diet and onset of supplementation, the mouse strains were different. While in heterogeneous mice GTE did not expand lifespan when supplemented from month 4⁽ Reference Strong, Miller and Astle ^{76 )} or 12⁽ Reference Spindler, Mote and Flegal ^{77 )}, in male C57BL/6 mice, supplementing GTE from age 13 months promoted longevity⁽ Reference Kitani, Osawa and Yokozawa ^{75 )}. In the case of the stilbene resveratrol lifespan extension also seems to depend on the genotype of the mice. While in C57BL/6 mice⁽ Reference Baur, Pearson and Price ^{96 )} resveratrol extended lifespan, in heterogeneous mice it did not⁽ Reference Miller, Harrison and Astle ^{97 )}. Furthermore, the impact of flavonoids on health may depend on the sex of the model organism. In male flies, EGCG extended lifespan although female flies did not benefit from EGCG supplementation⁽ Reference Wagner, Piegholdt and Rabe ^{60 )}. Therefore, the species and the genotype appear to be important for the responsiveness of an organism to polyphenol treatment and it seems that the more complex the organism, the more complex the effect of flavonoids on it is.

Challenges of deducing whether human subjects could benefit from flavonoid supplementation

Human flavonoid intake varies considerably. Vogiatzoglou et al. ⁽ Reference Vogiatzoglou, Mulligan and Lentjes ^{4 )} have reported that although the mean intake in Europe is 430 mg/d, median intake is much lower, 160 mg/d, meaning that some human subjects consume high amounts of flavonoids with their diets, while others consume very little. Thus, based on murine intake in the reviewed articles with a positive outcome for lifespan extension, human intake would be four to ten times lower compared with the amount mice are usually supplemented with (calculated as mg flavonoid/kg bodyweight) and the amounts supplemented in worm and fly studies are even higher. Interestingly, in human subjects, much lower doses than in animal trials benefitted health since in intervention trials with doses from 100 to 800 mg/d⁽ Reference Hooper, Kroon and Rimm ^{98 ,} Reference Egert, Bosy-Westphal and Seiberl ^{99 )}, positive effects on CVD-related parameters were reported. Furthermore, it would be interesting to study whether flavonoid consumption may abrogate the negative impact of high-fat and high-sugar diets on human health as it seems the case in flies and possibly mice. However, in some human trials, flavonoid consumption could not decrease CVD risk⁽ Reference Vogiatzoglou, Mulligan and Bhaniani ^{100 )} and cancer does not seem to be affected⁽ Reference Hui, Qi and Qianyong ^{101 )}. Unfortunately, randomised controlled or prospective trials studying human morbidity or mortality as an endpoint for flavonoid supplementation are scarce.