Updated bioavailability and 48h excretion profile of flavan-3-ols fromgreen tea in humans

LUCA CALANI1, DANIELE DEL RIO1, MARIA LUISA CALLEGARI2, LORENZO MORELLI3,

& FURIO BRIGHENTI1

1Human Nutrition Unit, The w2 Laboratory of Phytochemicals in Physiology, Department of Public Health, University of Parma,

Parma, Italy, 2Centro Ricerche Biotenologiche, Universita Cattolica del Sacro Cuore, Cremona, Italy, and 3Instituto di

Microbiologia Universita Cattolica del Sacro Cuore, Piacenza, Italy

AbstractGreen tea is a popular beverage, prepared with infusion of unfermented dried leaves of Camellia sinensis, and is one of the mostrelevant sources of polyphenolic compounds in the human diet. This study reports green tea flavan-3-ol absorption, metabolismand complete urinary excretion up to 48 h in 20 healthy volunteers. Urinary and tea samples were analysed by high-performanceliquid chromatography coupled with tandem mass spectrometry. Green tea contained monomeric flavan-3-ols andproanthocyanidins with a total polyphenol content of 728mmol. A total of 41 metabolites were identified in urines, all presentin conjugated forms. Among these, six colonic metabolites of green tea flavan-3-ols were identified for the first time aftergreen tea consumption in humans. The average 48 h bioavailability was close to 62%, major contributors being microbialmetabolites. Some volunteer showed a 100% absorption/excretion, whereas some others were unable to efficientlyabsorb/excrete this class of flavonoids. This suggests that colonic ring fission metabolism could be relevant in the putativebioactivity of green tea polyphenols.

Keywords: green tea flavan-3-ols, colon microbiota, bioavailability, urinary excretion, mass spectrometry, polyphenols

Introduction

Green tea is a popular beverage, prepared with

infusion of dried leaves of Camellia sinensis, without

fermentation steps. It is a relevant source of phenolic

compounds, especially monomeric flavan-3-ols (aka

catechins), which have been suggested to present

beneficial effects against cardiovascular diseases and

other ageing-related disorders, based on in vitro and

epidemiological studies (Cabrera et al. 2006; Zaveri

2006; Thangapazham et al. 2007; Kuriyama 2008;

Mandel et al. 2008). Several works have been

published on the bioavailability and metabolism of

green tea catechins in humans. After ingestion, a

relevant fraction of tea catechins undergoes extensive

metabolism by phase II enzymes, such as UDP-

glucuronosyltransferases (UGTs), sulphotransferases

(SULTs) and catechol-O-methyltransferase (COMT)

before and after absorption in the small intestine and

in liver, respectively (Lambert et al. 2007). Further

phase II conjugation has been reported to take place

in the large intestine and in kidneys (Tukey and

Strassburg 2000; Coughtrie and Johnston 2001;

Knights and Miners 2010). The formed metabolites

are likely to reach tissues through systemic circulation

before being excreted in the urine. However, a

substantial amount of the ingested flavan-3-ols is not

absorbed in the small intestine, thus reaching the large

intestine. Here, the host microbiota is able to break

down the flavonoidic skeleton generating several low-

molecular-weight metabolites with the characteristic

phenolic acid structure, namely 4-hydroxybenzoic acid,

3-methoxy-4-hydroxyphenylacetic acid, 3-(30-hydro-

xyphenyl)-3-hydroxypropionic acid and g-valerolactones

(Sang et al. 2008; Selma et al. 2009). These small

phenolics could undergo phase II metabolism locally

and/or get absorbed and reach the liver for further

enzymatic conjugation before entering the systemic

ISSN 0963-7486 print/ISSN 1465-3478 online q 2011 Informa UK, Ltd.

DOI: 10.3109/09637486.2011.640311

Correspondence: Daniele Del Rio, Human Nutrition Unit, The w2 Laboratory of Phytochemicals in Physiology, Department of Public Health,University of Parma, Via Volturno 39, 43125 Parma, Italy. Tel: þ 39 0521 903830. Fax: þ 39 0521 903832. E-mail: daniele.delrio@unipr.it

International Journal of Food Sciences and Nutrition,

August 2012; 63(5): 513–521

Int J

Foo

d Sc

i Nut

r D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y Y

ork

Uni

vers

ity L

ibra

ries

on

11/0

4/14

For

pers

onal

use

onl

y.

circulation and being excreted in urine in quantities

that largely exceed the metabolites generated by small

intestine absorption of the intact parent compound

(Crozier et al. 2010). Based on the most recent and

complete studies (Stalmach et al. 2009; Roowi et al.

2010), excretion of urinary phenolic metabolites was

210mmol after ingestion of 634mmol of flavan-3-ols,

corresponding to a 27% degradation of the ingested

dose, which, summed to an 8% excretion of

glucuronide, sulphate and methylated flavan-3-ols

originating from absorption in the small intestine,

accounts for 35% absorption. However, based on

our previous work (Del Rio et al. 2010b), microbic

valerolactone metabolites were excreted in quantities

equivalent to 36% of intake and were not included

in the previous estimate. When summed up, these

values give a total excretion of about 71% of intake.

However, this estimate is obviously an approximation

as it derives by adding results from different studies

dissimilar in designs and methodologies. Moreover,

this figure might be underestimated, being calculated

considering a 24 h urinary excretion, even if we have

previously reported that g-valerolactones are excreted

up to 2 days after intake of green tea (Del Rio et al.

2010b). Therefore, the aim of the present study was

to investigate green tea flavan-3-ol absorption by

measuring urinary excretion and metabolic profile

over 48 h by means of high-performance liquid

chromatography coupled with tandem mass spec-

trometry (HPLC–MS/MS).

Methods

Tea and chemicals

Five-hundred-millilitre bottles of ready-to-drink

(RTD) green tea were industrially made from Sri

Lankan tea leaves. The manufacturing process was

based on infusion in hot water, reproducing the

traditional tea preparation. The RTD tea is composed

of tea infusion (water, tea), sugar, peach juice,

dextrose, lemon juice, ascorbic acid and flavours. Pure

(2)-epicatechin, (2)-epigallocatechin, (2)-epigallo-

catechin-3-gallate and (2 )-epicatechin-3-gallate

standards were obtained from Sigma (St. Louis, MO,

USA), while procyanidin B2 was purchased from

Extrasynthese (Genay Cedex, France). All the solvents

and reagents were purchased from Carlo Erba

Reagenti (Milano, Italy).

Human feeding study

The feeding study was carried out on 20 healthy

human volunteers. Exclusion criteria included dia-

betes mellitus, cardiovascular events, chronic liver

diseases or nephropathies, cancer, organ failure and

use of antioxidant or vitamin supplements. The

volunteers were 25 ^ 3 years old (mean ^ SD), with

an average BMI of 22 ^ 3 kg/m2. The study protocol

was approved by the Ethics Committee for Human

Research of the University of Parma, and each

volunteer signed an informed consent before entering

the trial.

For 2 days prior to, and up to 48 h after the ingestion

of tea, the subjects followed a diet deprived of

flavonoids and phenolic compounds by avoiding fruit

and fruit juices, chocolate, nuts, vegetables, tea and

any kind of herbal tea, coffee, wine and dietary

antioxidant supplements. To check for compliance,

the volunteers were asked to complete a 4-day weighed

food record during the 2 days before and the 2 days

after the test drink of green tea. On the day of the test,

after an overnight fast, each subject drank 500 ml of

green tea. Urine was collected at time 0 (basal) and at

0–4, 4–7, 7–10, 10–24, 24–28, 28–34 and 34–48 h

collection periods after tea ingestion. The volume of

urine collected during each period was measured, and

three 5-ml samples are stored at 2808C until analysis.

Suitable aliquots of urine samples were filtered with

0.45-mm nylon filter (Waters, Milford, MA, USA) and

directly analysed by HPLC–MS/MS without further

processing.

HPLC–ESI–MS/MS analysis

Flavan-3-ols and their metabolites in tea and urine

were analysed using a Waters Alliance 2695 Separation

Module equipped with a Micromass Quattro Micro

API mass spectrometer fitted with an electrospray

interface (ESI; Waters). Separations were performed

using a Waters Atlantis dC18 3mm (2.1 £ 150 mm)

reverse phase column (Waters). For tea catechins, the

mobile phase, pumped at a flow rate of 0.17 ml/min,

was a 30-min linear gradient of 5–30% acetonitrile in

1% aqueous formic acid. The tuning of the mass

spectrometer was optimized by infusing a standard of

(2 )-epicatechin into the source along with 5%

acetonitrile in 1% aqueous formic acid, the initial

HPLC mobile phase, at a flow rate of 30ml/min. The

ESI source worked in negative ionization mode.

Source temperature was 1208C, desolvation tempera-

ture was 3508C, capillary voltage was 2.8 kV and cone

voltage was 35 V. The collision energy for MS/MS

identifications was set at 25 eV. For tea proanthocya-

nidins (PAs), the analytical conditions used were the

same as for catechins, with the exception of capillary

voltage set at 3.2 kV and collision energy set at 30 eV.

In all analyses, the desolvation gas (nitrogen) was

750 l/h, the cone gas (nitrogen) was 50 l/h and the

collision gas used was argon. Following HPLC

separation and MS/MS identification, catechins and

their metabolites were quantified using HPLC with

the MS operating in the selected ion recording (SIR)

mode, while to quantify green tea PAs, the spec-

trometer was operated in multiple reaction monitoring

(MRM) mode. Tea catechins were quantified using

calibration curves of the appropriate standard or

using the respective epimers (e.g. gallocatechin in

L. Calani et al.514

Int J

Foo

d Sc

i Nut

r D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y Y

ork

Uni

vers

ity L

ibra

ries

on

11/0

4/14

For

pers

onal

use

onl

y.

epigallocatechin equivalents), while green tea PAs

were quantified in procyanidin B2 equivalents. The

metabolites of epicatechin and epigallocatechin in

urine were quantified using epicatechin and epigallo-

catechin, respectively. Glucuronidated, methylated

and sulphated ring-fission metabolites of tea flavan-

3-ols (namely g-valerolactones) were quantified as

epicatechin equivalents.

Results

Analysis of tea

The flavan-3-ol content of 500 ml of RTD green tea

(n ¼ 5, each sample coming from a different pack, all

packs from the same lot) is reported in Table I together

with the MS/MS characteristics of each analysed

compound.

Identification of urinary flavan-3-ol metabolites

The HPLC–MS/MS analysis allowed the identifi-

cation of 41 different urinary flavan-3-ol metabolites,

all reported in Table II with their mass spectral

characteristics.

In detail, several metabolites of (epi)catechin (EC)

and (epi)gallocatechin (EGC) derived by the action of

intestinal or hepatic UGTs, SULTs and COMT were

identified by monitoring the loss of the conjugating

groups (i.e. glucuronic acid and sulphate) to give the

Table I. Mass spectral characteristics and content of monomeric and oligomeric flavan-3-ols in green tea.

Compound

Monomeric

units

[M 2 H]2

(m/z)

Qualifier ions

(m/z)

Content/500 ml

of RTD (mmol)*

(2)-Epigallocatechin – 305 – 188.3 ^ 16.6

(þ)-Gallocatechin – 305 – 35.7 ^ 4.1

(2)-Epicatechin – 289 – 49.1 ^ 5.5

(þ)-Catechin – 289 – 8.2 ^ 1.0

(2)-Epigallocatechin-3-O-gallate – 457 – 198.7 ^ 17.5

(þ)-Gallocatechin-3-O-gallate – 457 – 21.4 ^ 2.6

(2)-Epicatechin-3-O-gallate – 441 – 132.8 ^ 14.4

(þ)-Catechin-3-O-gallate – 441 – 5.0 ^ 0.7

Procyanidin dimers, B-type (epi)C ! (epi)C 577 289 42.3 ^ 5.4

Prodelphinidin dimers, B-type (epi)C ! (epi)GC 593 289 9.5 ^ 1.3

Prodelphinidin dimers, B-type (epi)GC ! (epi)GC 609 305 15.2 ^ 2.1

Procyanidin-gallate dimers, B-type (epi)C ! (epi)CG 729 289,407 8.2 ^ 1.8

Prodelphinidin-gallate dimers, B-type (epi)C ! (epi)GCG 745 289 6.4 ^ 1.6

Prodelphindin-gallate dimers, B-type (epi)GC ! (epi)GCG 761 169,305 6.8 ^ 1.3

Note: (epi)C, (epi)catechin; (epi)GC, (epi)gallocatechin; (epi)CG, (epi)catechin-gallate; (epi)GCG, (epi)gallocatechin-gallate; * Mean

values ^ SD, n ¼ 5; each sample coming from a different pack, all packs from the same lot.

Table II. MS/MS identification of urinary metabolites derived from green tea flavan-3-ols.

Metabolite [M 2 H]2 (m/z) Qualifier ions (m/z) Isomer numbers

Methyl-(epi)gallocatechin-sulphate glucuronide 575 495,399,319 2

Methyl-(epi)gallocatechin-sulphate 399 319 2

Methyl-(epi)gallocatechin-glucuronide 495 319 1

(Epi)gallocatechin-glucuronide 481 305 1

(Epi)gallocatechin-sulphate 385 305 3

(Epi)catechin-sulphate glucuronide 545 465,369,289 4

(Epi)catechin-sulphate 369 289 3

Methyl-(epi)catechin-sulphate 383 303 5

(Epi)catechin-glucuronide 465 289 1

M7-sulphate 271 191,147 1

M7-glucuronide 367 191 1

M60-glucuronide 383 207,163 1

M60-disulphate 367 287,207,163 1

M60-sulphate 287 207 1

M6-sulphate glucuronide 463 287,207 1

M6-glucuronide 383 207,163 2

M6-disulphate 367 287,207,163 1

Methyl-M6-glucuronide 397 221 1

M4-sulphate 303 223,179 2

Methyl-M4-sulphate 317 237 2

M4-glucuronide 399 223,179 2

Methyl-M4-glucuronide 413 237 3

Note: M7, hydroxyphenyl-g-valerolactone; M4, 5-(30,40,50-trihydroxyphenyl)-g-valerolactone; M6, 5-(40,50-dihydroxyphenyl)-g-valerolactone;

M60, 5-(30,50-dihydroxyphenyl)-g-valerolactone.

Updated bioavailability of flavan-3-ols 515

Int J

Foo

d Sc

i Nut

r D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y Y

ork

Uni

vers

ity L

ibra

ries

on

11/0

4/14

For

pers

onal

use

onl

y.

aglycone fragment ion (i.e. EC at m/z 289, EGC at m/z

305 and their methylated counterparts, at m/z 303 and

319, respectively), as previously reported by different

studies on urinary excretion of green tea flavan-3-ols

(Auger et al. 2008; Stalmach et al. 2009, 2010). The

ring-fission products of tea flavan-3-ols, 5-(30,40-

dihydroxyphenyl)-g-valerolactone (M6), 5-(30,50-

dihydroxyphenyl)-g-valerolactone (M60), 5-(30,40,50-

trihydroxyphenyl)-g-valerolactone (M4) and 5-

(hydroxyphenyl)-g-valerolactone (M7) of microbial

origin were solely detected in conjugated form, linked

to glucuronide, sulphate and methyl groups. The

identification criteria were the same used for EC and

EGC metabolites, monitoring the loss of conjugating

groups to generate the aglycones (i.e. M7 at m/z 191,

M4 at m/z 223, M6 and M60 at m/z 207 or the

methylated counterpart of M4 and M6 at m/z 237 and

221, respectively). The specific identification of M6

and M60 conjugates, which share the same molecular

weight, was obtained through the analysis of urine

after consumption of sources of EC and procyanidins,

but free of EGC or prodelphinidins, as M60 can derive

solely, like M4, from flavan-3-ols with three hydroxyl

groups on the B ring, unlike M6 which can derive

from all green tea flavan-3-ols (Sang et al. 2008). We

did not identify the exact position of the hydroxyl

group for M7, and it is therefore impossible to know if

this metabolite derived from the EC or the EGC

parent unit.

The urinary excretion profiles of the most relevant

EC and EGC phase II metabolites over a period of

48 h are reported in Figure 1, while Figure 2 shows the

excretion profiles related to the main ring-fission

metabolites of green tea flavan-3-ols.

Figure 1. Urinary excretion profiles of the main EC and EGC metabolites during 48 h. Data expressed as mean values with their standard

errors (n ¼ 20).

L. Calani et al.516

Int J

Foo

d Sc

i Nut

r D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y Y

ork

Uni

vers

ity L

ibra

ries

on

11/0

4/14

For

pers

onal

use

onl

y.

Quantitative analysis of urinary catabolites and their

excretion profile

For metabolite quantification, urine samples were

run in the HPLC–MS/MS operating in SIR mode.

This allowed a more accurate evaluation of flavan-3-ol

bioavailability (Del Rio et al. 2010b). The quantitative

excretions of flavan-3-ol metabolites in urine up to

48 h are reported in Table III.

As previously reported (Del Rio et al. 2010b;

Stalmach et al. 2010), among the parent flavan-3-ols

present in green tea, only EC and EGC were recovered

in urine, all in conjugated form due to the interaction

with human phase II enzymes. In our sample of

volunteers, EC and EGC showed a different metabolic

behaviour, as methyl-EC-sulphate (average excretion

equal to 2.51mmol) and its unmethylated counterpart

(1.34mmol) were the main urinary metabolites of

EC, followed by the glucuronide (1.28mmol) and the

sulphate glucuronide derivatives (0.43mmol). On

the contrary, the main urinary metabolites of EGC

were the glucuronide forms (7.35mmol) and their

methylated counterparts (4.67mmol), followed by

methyl-EGC-sulphate (1.97mmol), methyl-EGC-sul-

phate–glucuronide (0.86mmol) and EGC-sulphate

with 0.41mmol excreted on average.

Concerning ring-fission metabolites, urinary

excretion largely exceeded that of the flavan-3-ol

conjugates. Among these molecules, the main

g-valerolactone was by far the M60-disulphate, with

almost 163mmol excreted on average, followed by the

same conjugate of M6 (about 87.6mmol). As for

disulphates, the glucuronidated forms of M60 (about

34.4mmol) were higher than those of M6 (around

16.8 mmol). The M6-sulphate–glucuronide and

methyl-M6-glucuronide were the lesser phase II

metabolites of M6, and M60 was not recovered in

these forms. For M4, the methyl-sulphate derivatives

(about 54.7mmol) and its unmethylated counterpart

(about 27.7mmol) were the main urinary metabolites,

followed by the glucuronide (around 12.1mmol) and

the methyl-glucuronide (almost 2.7mmol). Finally,

also the newly observed M7 showed to preferably

interact with SULTs, with almost 19.7mmol of

monosulphate vs. only 6.6mmol of the glucuronide

conjugate.

Most of the EC and EGC conjugates reached their

excretive peak during the 0–4 and 4–7 h collection

periods. In detail, methyl-EC-sulphate, EC-sulphate

and the glucuronide forms of EC and EGC peaked

within the 4th hour from tea ingestion, whereas the

sulphate and glucuronide forms of methyl-EGC were

mainly excreted during the second collection time,

as well as EC-sulphate–glucuronide and methyl-

EGC-sulphate glucuronide. Only EGC-sulphate

showed an average excretive peak during the 7–10 h

collection period. Microbiota-derived g-valerolac-

tones showed a completely different excretive kinetic.

They are generated in the colon at a later stage

compared with catechin conjugates, and are therefore

absorbed and excreted with a delay of several hours,

with maximum urinary excretion at 24 h.

Finally, the bioavailability of green tea flavan-3-ols

was calculated as the ratio between the total metabolite

excretion (flavan-3-ols and microbial metabolites)

and the total intake of flavan-3-ols (catechins plus

PAs). The average flavan-3-ol bioavailability calcu-

lated from total metabolites recovered over 48 h

was 61.9 ^ 31.2% (mean ^ SD), with inter-subject

variations ranging from 17.5 to 100.0% among the

20 subjects.

Figure 2. Urinary excretion profiles of the main ring-fission metabolites during 48 h. Data expressed as mean values with their standard errors

(n ¼ 20).

Updated bioavailability of flavan-3-ols 517

Int J

Foo

d Sc

i Nut

r D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y Y

ork

Uni

vers

ity L

ibra

ries

on

11/0

4/14

For

pers

onal

use

onl

y.

Table

III.

Qu

an

tifi

cati

on

of

pri

nci

pal

flav

an

-3-o

lm

etabolite

sas

excr

eted

inu

rin

eat

dif

fere

nt

collec

tion

tim

es,

aft

erco

nsu

mpti

on

of

500

ml

of

RT

Dgre

ente

aby

20

hu

man

volu

nte

ers.

Met

ab

olite

0–

4h

4–

7h

7–

10

h10

–24

h24

–28

h28

–34

h34

–48

h

Met

hyl-

EG

C-s

ulp

hate

glu

curo

nid

e0.1

4^

0.0

90.3

0^

0.1

10.1

6^

0.1

20.1

8^

0.1

30.0

6^

0.1

80.0

1^

0.0

40.0

1^

0.0

4

Met

hyl-

EG

C-s

ulp

hate

0.5

8^

0.8

00.8

0^

0.6

30.2

8^

0.3

70.2

9^

0.4

70.0

2^

0.0

60.0

0^

0.0

1n

.d.

Met

hyl-

EG

C-g

lucu

ron

ide

1.6

1^

0.9

91.6

2^

1.0

00.6

1^

0.4

40.6

7^

0.6

00.1

1^

0.0

90.0

3^

0.0

40.0

1^

0.0

3

EG

C-g

lucu

ron

ide

3.7

6^

2.0

12.5

7^

1.5

40.6

1^

0.4

50.3

5^

0.2

80.0

5^

0.0

60.0

1^

0.0

2n

.d.

EG

C-s

ulp

hate

0.0

2^

0.0

40.1

2^

0.1

10.2

4^

0.3

20.0

4^

0.1

2n

.d.

n.d

.n

.d.

EC

-su

lphate

glu

curo

nid

e0.1

3^

0.0

80.1

8^

0.0

70.0

6^

0.0

50.0

5^

0.0

50.0

0^

0.0

1n

.d.

n.d

.

EC

-su

lphate

0.5

7^

0.6

80.4

5^

0.3

40.1

4^

0.1

60.1

6^

0.3

30.0

1^

0.0

30.0

0^

0.0

20.0

1^

0.0

4

Met

hyl-

EC

-su

lphate

0.9

5^

0.5

60.8

5^

0.4

70.3

0^

0.2

20.3

5^

0.1

90.0

3^

0.0

60.0

0^

0.0

20.0

2^

0.0

9

EC

-glu

curo

nid

e0.6

2^

0.3

90.4

5^

0.3

10.1

2^

0.0

70.0

9^

0.0

80.0

1^

0.0

10.0

0^

0.0

10.0

0^

0.0

1

M7-s

ulp

hate

0.1

2^

0.3

11.2

9^

2.3

92.2

9^

3.8

79.0

1^

11.2

64.5

7^

5.7

11.6

8^

2.2

20.7

3^

0.6

9

M7-g

lucu

ron

ide

0.0

3^

0.0

70.4

5^

0.5

81.0

3^

1.3

63.2

2^

3.4

41.4

1^

1.9

70.2

7^

0.2

90.1

8^

0.2

0

M60 -

glu

curo

nid

e0.4

2^

0.5

12.6

1^

3.3

44.4

4^

4.5

915.9

9^

11.6

37.1

7^

7.8

82.3

7^

1.6

21.3

7^

1.3

2

M60 -

dis

ulp

hate

0.5

3^

1.3

28.1

1^

9.9

515.4

1^

19.2

974.5

7^

54.5

741.4

1^

43.9

614.9

9^

8.1

17.9

0^

7.0

7

M6-s

ulp

hate

glu

curo

nid

e0.0

4^

0.0

90.3

7^

0.2

60.4

1^

0.3

60.9

3^

0.7

20.3

9^

0.6

00.0

7^

0.0

60.0

1^

0.0

3

M6-g

lucu

ron

ide

0.3

5^

0.5

44.0

4^

4.0

64.2

8^

3.0

86.7

7^

5.4

01.0

8^

1.5

10.1

5^

0.1

30.0

8^

0.1

3

M6-d

isu

lphate

3.8

4^

5.7

216.7

4^

13.6

116.5

4^

17.8

334.4

5^

25.3

911.3

1^

13.3

53.2

9^

2.5

01.4

2^

1.2

5

Met

hyl-

M6-g

lucu

ron

ide

0.1

5^

0.5

00.2

3^

0.1

80.2

5^

0.2

00.7

3^

0.5

60.2

9^

0.3

30.0

8^

0.0

50.0

8^

0.1

1

M4-s

ulp

hate

1.0

2^

1.7

56.1

0^

6.8

67.1

9^

10.3

312.6

2^

14.6

80.6

9^

0.9

40.0

7^

0.1

60.0

3^

0.1

2

Met

hyl-

M4-s

ulp

hate

1.2

7^

2.3

910.1

5^

11.4

215.1

8^

20.3

826.4

2^

30.3

81.5

0^

2.0

50.1

4^

0.2

60.0

2^

0.0

4

M4-g

lucu

ron

ide

0.2

2^

0.2

92.6

5^

3.3

03.3

8^

3.8

95.4

3^

5.8

70.3

3^

0.4

80.0

4^

0.1

00.0

7^

0.1

5

Met

hyl-

M4-g

lucu

ron

ide

0.0

2^

0.0

40.3

7^

0.4

40.6

0^

0.6

61.4

4^

1.5

10.1

5^

0.2

00.0

3^

0.0

80.0

5^

0.2

1

Note

:M

ean

valu

esin

mm

ol^

SD

,n¼

20.

n.d

.–

not

det

ecte

d.

L. Calani et al.518

Int J

Foo

d Sc

i Nut

r D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y Y

ork

Uni

vers

ity L

ibra

ries

on

11/0

4/14

For

pers

onal

use

onl

y.

Discussion

The RTD green tea used in this study is a rich source

of flavan-3-ols. In particular, the beverage provided

more than 639mmol of monomeric catechins (equal to

approximately 247 mg) per bottle (500 ml). Moreover,

the beverage contained about 88mmol of oligomeric

flavan-3-ols, also known as PAs. Although these

molecules had been already identified in green tea

(Lin et al. 2008; Kalili and de Villiers 2010), to our

knowledge this is the first work on bioavailability and

metabolism of tea polyphenols which takes into

account PAs. It has recently been reported that PAs

may be converted by the human faecal microbiota to

several low-molecular-weight metabolites, such as

phenyl-g-valerolactones (Duweler and Rohdewald

2000; Appeldoorn et al. 2009; Stoupi et al. 2010).

For this reason, the inclusion of PAs in the total flavan-

3-ol quantification is required to avoid underestima-

tion in the bioavailability calculation.

Several works investigated metabolism of green tea

flavan-3-ols in humans, focusing on phase II metab-

olites formed in the small intestine and liver and

estimating bioavailability values close to 8–9% (Auger

et al. 2008; Stalmach et al. 2009). This low

bioavailability value was demonstrated to increase to

39% when colon-derived valerolactones were con-

sidered (Del Rio et al. 2010b). In the present study, 41

different flavan-3-ol metabolites were recovered in

urine, accounting for phase II conjugates of both

parent catechins and microbial ring-fission products.

Several flavan-3-ol conjugates reached their excretion

peak later than previously reported (Del Rio et al.

2010b), this trend being more marked for methyl-

EGC-sulphate, its glucuronidated counterparts and

EC-sulphate-glucuronide, which peaked during the

4–7 h collection interval. Such delayed excretions

could be explained if we consider that these

metabolites derive primarily by hepatic conjugations

or from enterohepatic recirculation. Neither EGCG

nor ECG was identified in urine, in keeping with

previous studies on human metabolism of catechins

(Manach et al. 2005). As expected, PAs were not

detected in urine in their aglycone or conjugated forms

in the present study. Actually, most of the large-sized

flavan-3-ols (EGCG, ECG and PAs) are poorly

absorbable in the small intestine but are efficiently

converted to small-sized microbial metabolites in the

colon (Selma et al. 2009).

Four ring-fission metabolites of flavan-3-ols were

recovered in urine, namely M4, M60, M6 and M7. All

of them were present in their conjugated forms, the

aglycones being completely absent in all volunteers at

all the time points. They are known to be generated in

the colon at a later stage compared with catechin

conjugates, and are therefore absorbed and excreted

with a delay of several hours, with a maximum urinary

excretion at 24 h and quantifiable levels still present

at the last urine collection (48 h). Although all

ring-fission metabolites reached their excretion peak

around 24 h after ingestion, they were not equivalent

in their excretive kinetic. In fact, M6 conjugates were

generally excreted in larger amounts in the first

collection hours with respect to M60 phase II

metabolites, which largely exceeded M6 only after

the 10th hour from tea intake. A difference in

generation pathway into the colon could be a likely

explanation; in fact, M6 has a 30- and a 40-hydroxyl

groups and derives from the breakdown of EC, ECG

and procyanidins, whereas an additional dehydroxyla-

tion step at the 40-hydroxyl group on the B ring is

necessary to form M60 after the breakdown of EGC,

EGCG and prodelphinidins.

For the quantification of flavonoid metabolites,

because only a few are commercially available and they

are very complicated to synthesize, the analytical

approach used in the past almost invariably involved

treatment of samples with hydrolytic enzymes –

namely glucuronidases and sulphatases – followed

by the quantification of the released aglycones by

HPLC using either absorbance, fluorescence or

electrochemical detection (Lee et al. 2002; Chow

et al. 2005; Henning et al. 2005). More recently,

HPLC with MS/MS detection without recourse to

enzyme hydrolysis emerged as the ideal technology to

study polyphenol bioavailability (Mullen et al. 2006,

2008). Indeed, MS/MS revealers allow the direct

identification of several phase II conjugates without

the need of previous hydrolysis. However, in the

absence of standards, it is not possible to distinguish

between isomers and to clearly identify the position of

conjugating groups on the flavonoid skeleton. For

example, a metabolite such as EC-30-glucuronide can

be partially identified as an EC-glucuronide on the

basis of its MS fragmentation pattern (Del Rio et al.

2010a). Despite this limitation, the use of tandem MS

allows the analysis of low nanomole concentrations

and provides structural information on analytes of

interest that is not obtainable with other detectors.

Quantification of identified metabolites by MS is

better performed using SIR and, as it can rarely be

based on calibration curves of exact standards, it is

generally carried out using the aglycone as a

comparable external standard. However, the slopes

of the dose-response curves of the two compounds are

not necessarily identical, and this might represent a

potential source of error in the quantitative analysis.

Nevertheless, the glucuronidase/sulphatase sources

used in sample preparation in previous works are not

constant in the enzyme titre and activity and there can

be substantial batch-to-batch variation in their

specificity (Donovan et al. 2006). Because the

enzymatic deconjugation approach is generally linked

to unavailability of MS facilities, there is a large lack of

information about the efficiency by which the enzymes

hydrolyse the individual metabolites and release the

aglycone. This introduces a varying, unmeasured error

factor. The quantitative estimates based on the use of

Updated bioavailability of flavan-3-ols 519

Int J

Foo

d Sc

i Nut

r D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y Y

ork

Uni

vers

ity L

ibra

ries

on

11/0

4/14

For

pers

onal

use

onl

y.

glucuronidase/sulphatase preparations are, therefore,

probably no better in accuracy than those based on

HPLC–MRM/SIR and, additionally, do not give

information about the specific conjugated metabolites,

leaving the putative bioactive forms unknown. More-

over, the detection of aglycones after deconjugating

processes implies the exclusion of methoxy derivatives,

which constitutes a relevant fraction of the total

urinary metabolites (up to 15% in this study).

Furthermore, in the present work, the use of

HPLC–MS/MS allowed the identification of three

novel conjugates of ring-fission metabolites, namely

methyl-M6-glucuronide, M7-sulphate and M7-glu-

curonide, which were previously identified in human

urines only after almond skin consumption (Llorach

et al. 2010). On the contrary, methyl-M4-glucuronide,

M6- and M60-disulphates, identified in our study, had

never been observed before.

The average bioavailability of tea flavan-3-ols

observed in this study was close to 62%, higher than

all the individual studies previously published in the

literature (Li et al. 2000; Auger et al. 2008; Stalmach

et al. 2009; Del Rio et al. 2010b). The main reasons

for this higher bioavailability value are the urinary

collection period, which lasted 2 days after tea

ingestion allowing an almost complete excretion of

flavan-3-ol metabolites, and the identification and

quantification of novel molecules derived by microbial

and subsequent phase II modification of catechins.

In fact, the excretion of previously identified meta-

bolites during the first 24 h accounted for 44% of

the ingested dose of catechins from tea, a value that

is in line with that calculated in our previous work

(Del Rio et al. 2010b).

The bioavailability values greatly varied among

volunteers. Actually, in five subjects, the urinary

excretion of flavan-3-ol metabolites was equivalent to

less than 30% of the ingested dose, whereas four

volunteers attained an absorption that reached 100%.

The main factor responsible for the inter-individual

variability was the g-valerolactones, as the total

excretion of M60-disulphate ranged from 42.9 to

336.2mmol, M6-disulphate ranged from 5.6 to

189.3mmol, methyl-M4-sulphate ranged from 0.5

to 149.2mmol and M7-sulphate ranged from 0.2 to

70.6mmol. This remarkable variability is in good

agreement with the previous works. Roowi and

colleagues (Roowi et al. 2010) incubated flavan-3-ols

with human faecal slurries in vitro to evaluate the

newly formed metabolites and found that M6 varied

more than an order of magnitude among three

faecal donors. Moreover, in a feeding study carried

out with 1.2 g of green tea extract ingested by five

human male volunteers, the urinary excretion of

valerolactones ranged from 1.6 to 90.3mmol and from

2.9 to 43.2mmol for M4 and M6, respectively

(Li et al. 2000).

Finally, the fact that blood samples have not been

analysed for their content in flavan-3-ol metabolites

could be considered the main limitation of this study.

The absence of standardized procedures for the

extraction of these molecules from plasma and the

concomitant absence of adequate standards to

calculate the yield of extraction were the main reasons

to limit our work to urine samples.

Conclusion

Monitoring urinary flavan-3-ol metabolite excretion

up to 48 h allowed us to provide better estimates of

the bioavailability of green tea flavan-3-ols, which

is largely attributable to the microbiota-derived

compounds. Some of the volunteers showed a 100%

recovery in urines, indicating complete absorption,

whereas some others were unable to efficiently

absorb/excrete this class of flavonoids. This suggests

that colonic ring-fission metabolism could be a

relevant factor in the bioactivity of green tea

polyphenols and should be taken in careful consi-

deration. A detailed characterization of the colonic

microbial population of each individual volunteer

is presently ongoing to try to unravel the possible

link between specific microbiota and the ability

to efficiently metabolize and absorb green tea

flavan-3-ols.

Acknowledgement

The authors would like to thank all the volunteers who

participated in the present study.

Declaration of interest: The authors report no

conflicts of interest. The authors alone are responsible

for the content and writing of the paper.

References

Appeldoorn MM, Vincken JP, Aura AM, Hollman PC, Gruppen H.

2009. Procyanidin dimers are metabolized by human microbiota

with 2-(3,4-dihydroxyphenyl)acetic acid and 5-(3,4-dihydroxy-

phenyl)-gamma-valerolactone as the major metabolites. J Agric

Food Chem 57(3):1084–1092.

Auger C, Mullen W, Hara Y, Crozier A. 2008. Bioavailability of

polyphenon E flavan-3-ols in humans with an ileostomy. J Nutr

138(8):1535S–1542S.

Cabrera C, Artacho R, Gimenez R. 2006. Beneficial effects of green

tea – a review. J Am Coll Nutr 25(2):79–99. Review.

Chow HH, Hakim IA, Vining DR, Crowell JA, Ranger-Moore J,

Chew WM, et al. 2005. Effects of dosing condition on the oral

bioavailability of green tea catechins after single-dose adminis-

tration of Polyphenon E in healthy individuals. Clin Cancer Res

11(12):4627–4633.

Coughtrie MW, Johnston LE. 2001. Interactions between dietary

chemicals and human sulfotransferases-molecular mechanisms

and clinical significance. Drug Metab Dispos 29(4 Pt 2):

522–528.

Crozier A, Del Rio D, Clifford MN. 2010. Bioavailability of dietary

flavonoids and phenolic compounds. Mol Aspects Med 31(6):

446–467.

Del Rio D, Borges G, Crozier A. 2010a. Berry flavonoids and

phenolics: bioavailability and evidence of protective effects. Br J

Nutr 104(Suppl 3):S67–S90.

L. Calani et al.520

Int J

Foo

d Sc

i Nut

r D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y Y

ork

Uni

vers

ity L

ibra

ries

on

11/0

4/14

For

pers

onal

use

onl

y.

Del Rio D, Calani L, Cordero C, Salvatore S, Pellegrini N, Brighenti

F. 2010b. Bioavailability and catabolism of green tea flavan-3-ols

in humans. Nutrition 26(11–12):1110–1116.

Donovan JL, Manach C, Fauks RM, Kroon PA. 2006. Absorption

and metabolism of dietary plant secondary metabolites. In:

Crozier A, Clifford MN, Ashihara H, editors. Plant secondary

metabolites: occurrence, structure and role in the human diet.

Oxford: Blackwell Publishing. p 303–351.

Duweler KG, Rohdewald P. 2000. Urinary metabolites of French

maritime pine bark extract in humans. Pharmazie 55(5):364–368.

Henning SM, Niu Y, Liu Y, Lee NH, Hara Y, Thames GD, et al.

2005. Bioavailability and antioxidant effect of epigallocatechin

gallate administered in purified form versus as green tea extract

in healthy individuals. J Nutr Biochem 16(10):610–616.

Kalili KM, de Villiers A. 2010. Off-line comprehensive two-

dimensional hydrophilic interaction £ reversed phase liquid

chromatographic analysis of green tea phenolics. J Sep Sci

33(6–7):853–863.

Knights KM, Miners JO. 2010. Renal UDP-glucuronosyltrans-

ferases and the glucuronidation of xenobiotics and endogenous

mediators. Drug Metab Rev 42(1):63–73.

Kuriyama S. 2008. The relation between green tea consumption and

cardiovascular disease as evidenced by epidemiological studies.

J Nutr 138(8):1548S–1553S, Review.

Lambert JD, Sang S, Yang CS. 2007. Biotransformation of green tea

polyphenols and the biological activities of those metabolites.

Mol Pharm 4(6):819–825.

Lee MJ, Maliakal P, Chen L, Meng X, Bondoc FY, Prabhu S, et al.

2002. Pharmacokinetics of tea catechins after ingestion of green

tea and (2)-epigallocatechin-3-gallate by humans: formation of

different metabolites and individual variability. Cancer Epide-

miol Biomarkers Prev 11(10 Pt 1):1025–1032.

Li C, Lee MJ, Sheng S, Meng X, Prabhu S, Winnik B, et al. 2000.

Structural identification of two metabolites of catechins and their

kinetics in human urine and blood after tea ingestion. Chem Res

Toxicol 13(3):177–184.

Lin LZ, Chen P, Harnly JM. 2008. New phenolic components and

chromatographic profiles of green and fermented teas. J Agric

Food Chem 56(17):8130–8140.

Llorach R, Garrido I, Monagas M, Urpi-Sarda M, Tulipani S,

Bartolome B, Andres-Lacueva C. 2010. Metabolomics study of

human urinary metabolome modifications after intake of almond

(Prunus dulcis (Mill.) D.A. Webb) skin polyphenols. J Proteome

Res 9(11):5859–5867.

Manach C, Williamson G, Morand C, Scalbert A, Remesy C. 2005.

Bioavailability and bioefficacy of polyphenols in humans.

I. Review of 97 bioavailability studies. Am J Clin Nutr

81(1 Suppl):230S–242S.

Mandel SA, Amit T, Kalfon L, Reznichenko L, Youdim MB. 2008.

Targeting multiple neurodegenerative diseases etiologies with

multimodal-acting green tea catechins. J Nutr 138(8):

1578S–1583S.

Mullen W, Edwards CA, Crozier A. 2006. Absorption, excretion

and metabolite profiling of methyl-, glucuronyl-, glucosyl- and

sulpho-conjugates of quercetin in human plasma and urine after

ingestion of onions. Br J Nutr 96(1):107–116.

Mullen W, Archeveque MA, Edwards CA, Matsumoto H, Crozier

A. 2008. Bioavailability and metabolism of orange juice

flavanones in humans: impact of a full-fat yogurt. J Agric Food

Chem 56(23):11157–11164.

Roowi S, Stalmach A, Mullen W, Lean ME, Edwards CA, Crozier A.

2010. Green tea flavan-3-ols: colonic degradation and urinary

excretion of catabolites by humans. J Agric Food Chem 58(2):

1296–1304.

Sang S, Lee MJ, Yang I, Buckley B, Yang CS. 2008. Human urinary

metabolite profile of tea polyphenols analyzed by liquid

chromatography/electrospray ionization tandem mass spec-

trometry with data-dependent acquisition. Rapid Commun

Mass Spectrom 22(10):1567–1578.

Selma MV, Espın JC, Tomas-Barberan FA. 2009. Interaction

between phenolics and gut microbiota: role in human health. J

Agric Food Chem 57(15):6485–6501.

Stalmach A, Troufflard S, Serafini M, Crozier A. 2009. Absorption,

metabolism and excretion of Choladi green tea flavan-3-ols by

humans. Mol Nutr Food Res 53(Suppl 1):S44–S53.

Stalmach A, Mullen W, Steiling H, Williamson G, Lean ME, Crozier

A. 2010. Absorption, metabolism, and excretion of green tea

flavan-3-ols in humans with an ileostomy. Mol Nutr Food Res

54(3):323–334.

Stoupi S, Williamson G, Drynan JW, Barron D, Clifford MN. 2010.

A comparison of the in vitro biotransformation of (2)-

epicatechin and procyanidin B2 by human faecal microbiota.

Mol Nutr Food Res 54(6):747–759.

Thangapazham RL, Singh AK, Sharma A, Warren J, Gaddipati JP,

Maheshwari RK. 2007. Green tea polyphenols and its

constituent epigallocatechin gallate inhibits proliferation of

human breast cancer cells in vitro and in vivo. Cancer Lett

245(1–2):232–241.

Tukey RH, Strassburg CP. 2000. Human UDP-glucuronosyl-

transferases: metabolism, expression, and disease. Annu Rev

Pharmacol Toxicol 40:581–616.

Zaveri NT. 2006. Green tea and its polyphenolic catechins:

medicinal uses in cancer and noncancer applications. Life Sci

78(18):2073–2080. Epub 2006 Jan 30. Review.

Updated bioavailability of flavan-3-ols 521

Int J

Foo

d Sc

i Nut

r D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y Y

ork

Uni

vers

ity L

ibra

ries

on

11/0

4/14

For

pers

onal

use

onl

y.