Phytomedicine 23 (2016) 1365–1374

Contents lists available at ScienceDirect

Phytomedicine

journal homepage: www.elsevier.com/locate/phymed

Anti-stress effects of drinking green tea with lowered caffeine and

enriched theanine, epigallocatechin and arginine on psychosocial

stress induced adrenal hypertrophy in mice

Keiko Unno

a , c , ∗, Ayane Hara

a , Aimi Nakagawa

a , Kazuaki Iguchi a , Megumi Ohshio

b , Akio Morita

b , Yoriyuki Nakamura

c

a Department of Neurophysiology, School of Pharmaceutical Sciences, University of Shizuoka, Shizuoka, Japan b Department of Functional Plant Physiology, Faculty of Agriculture, Shizuoka University, Shizuoka, Japan c Tea Science Center, Graduate Division of Nutritional and Environmental Sciences, University of Shizuoka, Shizuoka, Japan

a r t i c l e i n f o

Article history:

Received 4 January 2016

Revised 13 July 2016

Accepted 19 July 2016

Keywords:

Anti-stress effect

Arginine

Caffeine

Epigallocatechin

Green tea

Theanine

a b s t r a c t

Background: Theanine, an amino acid in tea, has significant anti-stress effects on animals and humans.

However, the anti-stress effects of drinking green tea have not yet been elucidated.

Hypothesis/purpose: The present study aimed to explore anti-stress effects of green tea and roles of tea

components in a mouse model of psychosocial stress.

Study design: We examined anti-stress effects of three types of green teas, theanine-rich “Gyokuro”, stan-

dard “Sencha”, and Sencha with lowered caffeine (low-caffeine green tea). Furthermore, the roles of tea

components such as caffeine, catechins, and other amino acids in anti-stress effects were examined.

Methods: To prepare low-caffeine green tea, plucked new tea leaves were treated with a hot-water spray.

Mice were psychosocially stressed from a conflict among male mice under confrontational housing. Mice

consumed each tea that was eluted with room temperature water ad libitum . As a marker for the stress

response, adrenal hypertrophy was compared with mice that ingested water.

Results: Caffeine was significantly lowered by spraying hot-water on tea leaves. While epigallocatechin

gallate (EGCG) is the main catechin in tea leaves, epigallocatechin (EGC) was mainly infused into water

at room temperature. Adrenal hypertrophy was significantly suppressed in mice that ingested theanine-

rich and low-caffeine green tea that were eluted with water at room temperature. Caffeine and EGCG

suppressed the anti-stress effects of theanine while EGC and arginine (Arg) retained these effects.

Conclusion: These results suggest that drinking green tea exhibits anti-stress effects, where theanine,

EGC and Arg cooperatively abolish the counter-effect of caffeine and EGCG on psychosocial stress induced

adrenal hypertrophy in mice.

© 2016 Elsevier GmbH. All rights reserved.

I

u

s

a

t

(

e

G

S

8

t

t

c

E

(

g

o

h

0

ntroduction

Theanine (L-theanine), the major amino acid and a sweet

mami component of green tea ( Camellia sinensis (L.) Kuntze ), has

ignificant anti-stress effects on animals and humans ( Kimura et

l., 2007; Unno et al., 2011, 2013a, 2013b ). In those studies, purified

heanine had been used. However, whether simply drinking green

Abbreviations: Ala, alanine; Arg, arginine; Asn, asparagine; Asp, aspartic acid;

+ )C, ( + )-catechin; EC, ( −)-epicatechin; ECG, ( −)-epicatechin gallate; EGC, ( −)-

pigallocatechin; EGCG, ( −)-epigallocatechin gallate; GABA, γ -amino butyric acid;

lu, glutamate; Gln, glutamine; Ser, serine. ∗ Correspondence author at: University of Shizuoka, School of Pharmaceutical

ciences, Department of Neurophysiology, 52-1 Yada, Suruga-ku, Shizuoka 422-

526, Japan. Fax: + 81 54 264 5909.

E-mail address: unno@u-shizuoka-ken.ac.jp (K. Unno).

t

t

a

o

s

f

h

ttp://dx.doi.org/10.1016/j.phymed.2016.07.006

944-7113/© 2016 Elsevier GmbH. All rights reserved.

ea exerts similar anti-stress effects is not known. We have shown

hat the anti-stress effect of theanine was blocked by caffeine and

atechins, two other main components of tea ( Unno et al., 2013a ).

ach tea component also provides a distinct taste such as umami

theanine and other amino acids), bitterness (caffeine), and astrin-

ency (catechins), which all together determine the overall taste

f a specific type of green tea. It is therefore important to know

he relationship between the biological function and taste of green

ea based on the balance of theanine, caffeine, catechins and other

mino acids.

The relative amounts of these tea components vary depending

n cultivation conditions such as the harvest season of tea leaves,

unshine conditions, breeds of tea plant, and geography. All these

actors affect the taste and grade of specific green teas. Gyokuro, a

igh-grade green tea, is prepared by placing in the shade for about

1366 K. Unno et al. / Phytomedicine 23 (2016) 1365–1374

f

c

t

a

s

w

(

5

1

(

5

t

1

A

b

w

o

e

(

p

(

l

i

u

m

T

3

t

i

p

i

(

o

a

G

c

s

s

1

6

e

i

g

(

t

M

c

a

F

d

A

J

c

w

p

three weeks before being harvested. This process increases both

theanine and caffeine in tea leaves ( Sasaoka et al., 1964; Konishi et

al., 1972 ). Gyokuro is thus umami-rich, but as a consequence, very

expensive. To fully infuse theanine, an umami component, from

Gyokuro, it is further necessary to steep in low-temperature water

(40–60 °C), whereas at a high temperature ( ∼90 °C), catechins and

caffeine are preferentially infused ( Shimotoku et al., 1982; Monobe

et al., 2010 ).

Sencha, a more commonly consumed “popular” green tea, is

grown under full sun, which results in reduced theanine and in-

creased catechins compared to Gyokuro. Furthermore, a grade of

Sencha green tea is generally distinguished by the amount of

amino acids, although a standard value has not been defined. In

this experiment, a higher grade Sencha was used to prepare Sen-

cha with lowered caffeine content (low-caffeine Sencha). A middle

grade Sencha was used as the standard. In this study, we propose a

practical way of maximizing the anti-stress effect of Sencha green

tea as a daily consumable drink by enriching theanine and lower-

ing caffeine and catechin through the temperature-sensitive kinet-

ics of water elution of each tea component. The anti-stress effect

of these three types of green tea was examined.

In the present study, the anti-stress effect of drinking green

tea was evaluated using our unique mouse model of psychosocial

stress evoked by confrontational housing. We previously demon-

strated that adrenal hypertrophy is a suitable marker for the stress

response ( Unno et al., 2013 a ). That is, two male mice were housed

in the same cage separated with a partition to establish a territo-

rial imperative. Then, the partition was removed and mice were

co-housed confrontationally. As a marker for the stress response,

changes in the adrenal gland were studied and compared with

group-housed control mice. Significant adrenal hypertrophy, which

was observed in mice under confrontational housing, had devel-

oped in 24 h and persisted for at least 1 week. The cell size in the

zona fasciculata of the adrenal gland, from which glucocorticoid is

mainly secreted, increased in mice under confrontational housing,

which was accompanied by a flat diurnal rhythm of corticosterone

and ACTH in blood. In addition, consumption of paroxetin, an anti-

depressant drug, significantly suppressed adrenal hypertrophy of

mice under confrontational housing. Therefore, adrenal hypertro-

phy is suitable for evaluating the anti-stress effect.

The present study examined whether the ingestion of green

tea prevents and relieves psychosocial stress. Furthermore, individ-

ual and combined effects of each tea component were examined

and the best balance among theanine, caffeine, catechins and other

amino acids was explored.

Materials and methods

Preparation of low-caffeine Sencha green tea

Plucked new tea leaves were treated with a hot-water spray at

95 °C as described previously ( Maeda-Yamamoto et al., 2007 ). To

compare the reduction of caffeine, tea leaves were treated with

heated water spray for 180 s and 280 s. After centrifugal dehydra-

tion, green tea was prepared through a standard manufacturing

process namely rolling and drying. We termed this low-caffeine

Sencha green tea ( Fig. 1 ). Tea components in both dried tea leaves

and eluates were measured by HPLC.

Measurement of tea components by HPLC

The chemical composition of dried tea leaves are generally mea-

sured to compare the grade and taste of green teas. However, in

this study, we focused on the biological functions of green tea

components in an elution steeped at room temperature. The tem-

perature was selected to reduce the elution of catechins and caf-

eine ( Shimotoku et al., 1982; Monobe et al., 2010 ). Therefore, tea

omponents in both dried tea leaves as described next and eluates

hat are described in Section (Green tea infusion at room temper-

ture and measurement of tea components by HPLC) were mea-

ured by HPLC.

The amount of caffeine, catechins and amino acids in tea leaves

as measured according to a method described in Horie et al.

2002) . In brief, powdered tea leaves (25 mg) were shaken in

ml of an acetonitrile–water solution (1:1 v/v) at 180 rpm for

h. Catechins and caffeine in the filtrate were measured by HPLC

SCL-10Avp, Shimadzu, Japan; Develosil packed column ODS-HG-

, 150 × 4.6 mm, Nomura Chemical Co. Ltd., Japan). Gradient elu-

ion from the column was performed at 40 °C at a flow rate of

.0 ml/min using a two-gradient elution system with mobile phases

(acetonitrile/water/phosphoric acid (10:400:1, v/v/v)) and B (mo-

ile phase A/methanol (10 0 0:50 0, v/v)). Catechins and caffeine

ere measured at 280 nm.

Next, free amino acids in tea leaves were measured using an-

ther elution condition. In brief, 10 mg of powdered tea leaves was

xtracted with 5 ml of water after adding polyvinylpolypyrrolidone

10 mg) to remove polyphenols. Alanine (Ala), arginine (Arg), as-

artic acid (Asp), asparagine (Asn), glutamic acid (Glu), glutamine

Gln), serine (Ser) and theanine, major amino acids in the tea

eaves, were measured by HPLC as described above using homoser-

ne as an internal standard ( Goto et al., 1993 ). Elution from the col-

mn was performed at 40 °C at a flow rate of 1.0 ml/min using the

obile phase of acetonitrile and 5 mM citric acid buffer (pH 6.0).

hese amino acids were detected at an excitation wavelength of

40 nm and at 450 nm of emission wavelength (RF-535 UV detec-

or, Shimadzu, Japan).

The HPLC profiles of standard catechins and caffeine, and those

n Sencha were shown in Fig. S1a and b, respectively. The HPLC

rofiles of standard amino acids and those in Sencha were shown

n Fig. S1c and d, respectively. The relative standard deviation

RSD%) of precision and repeatability were < 5.0%. The recoveries

f catechins, caffeine, and free amino acids were 99 ± 4%, 98 ± 4%,

nd 98 ± 3%, respectively.

reen tea infusion at room temperature and measurement of tea

omponents by HPLC

Three types of drinking green tea were prepared for mouse

tudies: theanine-rich Gyokuro, low-caffeine Sencha green tea, and

tandard Sencha green tea. Green tea leaves (3 g) were added to

l of tap water at room temperature and filtered after stirring for

min ( Fig. 2 ). Amino acids and non-gallate catechins such as ( −)-

pigallocatechin (EGC) and ( −)-epicatechin (EC) are infused easily

nto water at room temperature. On the other hand, caffeine and

allate catechins such as ( −)-epigallocatechin gallate (EGCG) and

−)-epicatechin gallate (ECG) are hardly soluble in room tempera-

ure water and are retained in tea leaves ( Shimotoku et al., 1982;

onobe et al., 2010 ). Each infusion was used to measure the tea

omponents by HPLC and for animal studies.

Catechins and caffeine in the infusions were measured by HPLC

s described in Section (Measurement of tea components by HPLC).

ree amino acids in the infusions were also measured by HPLC as

escribed in Section (Measurement of tea components by HPLC).

nimals, housing conditions for stress experiment

Male ddY mice (Slc: ddY, 4 weeks old) were purchased from

apan SLC Co. Ltd (Shizuoka, Japan) and kept in a conventional

ondition in a temperature- and humidity-controlled environment

ith a 12–12 h light–dark cycle (light period, 08:0 0–20:0 0 h; tem-

erature, 23 ± 1 °C; relative humidity, 55 ± 5%). Mice were fed with

K. Unno et al. / Phytomedicine 23 (2016) 1365–1374 1367

Caffeine is eluted

Sprayed hot-water shower at 95˚C

Low-caffeine Sencha

green tea

Centrifugaldehydration

Rolling

Drying

Fig. 1. Preparation of low-caffeine Sencha green tea. To reduce caffeine preferentially in tea leaves, a hot-water shower was sprayed onto new leaves of green tea at 95 °C for 180–280 s. Then, tea leaves that had been centrifuged to remove water were treated through a standard manufacturing process, which involved rolling and drying.

a

t

o

n

l

l

c

t

f

i

o

i

t

2

f

I

s

m

8

1

a

s

a

t

<

I

t

i

u

L

L

s

C

c

T

e

g

S

f

t

w

S

i

R

P

S

b

S

f

a

f

t

a

i

d

1

n

s

t

s

c

normal diet (CE-2; Clea Co. Ltd, Tokyo, Japan) and water ad libi-

um . All experimental protocols were approved by the University

f Shizuoka Laboratory Animal Care Advisory Committee (approval

o. 136068 and 166197) and were in accordance with the guide-

ines of the US National Institutes of Health for the care and use of

aboratory animals.

Mice at 4 weeks of age were housed in a group of six in a

age for 5 days for adaptation. The mice were then divided into

wo groups: confrontationally-housed and group-housed. For con-

rontational housing, a standard polycarbonate cage was divided

nto two identical subunits by a stainless-steel partition as previ-

usly described ( Unno et al, 2013a ). In brief, two mice were housed

n a partitioned cage for 1 week (single housing); then the parti-

ion was removed to expose the mice to confrontational stress for

4 h (confrontational housing). Each cage was placed in a Styro-

oam box in order to avoid social contact between cages.

ngestion of green tea in mice

The effects of theanine-rich Gyokuro, low-caffeine Sencha, and

tandard Sencha were examined in eight groups of mice (6

ice/group, total n = 48). Mice consumed each tea ad libitum for

days (single housing for 7 days and confrontational housing for

day) ( Fig. 2 ). Mice of group housing also consumed each tea

d libitum for 8 days. The drinking volume of each tea was mea-

ured. At the end of the experimental period, mice were sacrificed

nd adrenal glands were weighed. RSD of precision of weighting of

he adrenals within groups (confrontation and group housing) was

10.0%.

ngestion of individual and combined tea components in mice

The effects of individual and combined tea components such as

heanine, caffeine, catechins and other amino acids, were examined

n 34 groups of mice (4 mice/group, total n = 136). Tea components

sed were as follows: l -theanine (Suntheanine; Taiyo Kagaku Co.

td., Yokkaichi, Japan), EGCG (Sunphenon EGCg, Taiyo Kagaku Co.

td.), EGC (Sunphenon EGC, Taiyo Kagaku Co. Ltd.), ECG (Tokyo Ka-

ei Co. Ltd., Tokyo, Japan), Arg, Glu, and Gln (Wako Pure Chemical

o. Ltd., Osaka, Japan). Mice consumed individual or combined tea

omponents in tap water ad libitum for 8 days as described above.

he drinking volume of each tea component was measured. At the

nd of the experimental period, mice were sacrificed and adrenal

lands were weighed.

tatistical analysis

Data are expressed as mean ± SD. Statistical analyses were per-

ormed using one-way analysis of variance (ANOVA) followed by

he Bonferroni’s post hoc test for multiple comparisons. ANOVA

as assessed using a statistical analysis program (StatPlus, Analyst-

oft Inc., online version). Differences were considered to be signif-

cant at p < 0.05.

esults

reparation of low-caffeine Sencha green tea

First, we optimized a method to lower the level of caffeine in

encha tea leaves, while retaining theanine and other components,

y using a hot-water treatment ( Maeda-Yamamoto et al, 2007 ).

praying a heated-water shower at 95 °C preferentially eluted caf-

eine from plucked new tea shoots over 90 s ( Maeda-Yamamoto et

l, 2007 ). We treated Sencha tea leaves with a hot-water shower

or 180 or 280 s ( Fig. 1 ). As shown in Table 1 , caffeine decreased

o 1/4–1/5 of the level of non-treated tea leaves, whereas the

mount of catechins such as EGCG and EGC did not change signif-

cantly. EGCG was the most abundant followed by EGC. Theanine

ecreased slightly to 93% and 83% after hot-water treatment for

80 s and 280 s, respectively. The amount of other amino acids did

ot change. Taken together, treatment with hot water for 180 s is

ufficient to significantly lower caffeine without altering the other

ea components. We used Sencha leaves treated with a hot-water

pray for 180 s to create “low-caffeine Sencha” to measure the tea

omponents and for animal studies.

1368 K. Unno et al. / Phytomedicine 23 (2016) 1365–1374

Fig. 2. Tea leaves were eluted with water at room temperature. Tea leaves (3 g) were stirred in water (1 l) at room temperature for 6 min. The infusion was used to measure

tea components (catechins, amino acids, and caffeine) by HPLC and for animal studies to evaluate anti-stress effects.

E

T

t

S

f

A

t

r

f

a

Green tea infusions at room temperature

To reduce the elution of caffeine and gallate catechins, each

type of tea leaves (3 g) was steeped in water (1 l) at room temper-

ature by stirring for 6 min ( Fig. 2 ). The amounts of caffeine, cat-

echins and theanine in theanine-rich Gyokuro, low-caffeine Sen-

cha and standard Sencha green tea were compared ( Table 2 ). The

reduction of caffeine in low-caffeine Sencha was significant com-

pared to Gyokuro ( > 170-fold) and standard tea ( > 70-fold) ( Table

2 ). The most common catechin was non-gallate catechins such as

GC and EC, while EGCG was almost absent in each tea elution.

he total amount of catechins was highest in the standard tea. Al-

hough the amount of theanine in the infusion from low-caffeine

encha is half of that from Gyokuro, it is 2.5-fold higher than that

rom standard Sencha ( Table 2 ). Other amino acids such as Glu and

rg were slightly higher in low-caffeine green tea than in other

eas ( Table 2 ). Collectively, low-caffeine Sencha steeped in water at

oom temperature is characterized by a negligible amount of caf-

eine, significantly reduced gallate catechins and enriched theanine

nd other amino acids.

K. Unno et al. / Phytomedicine 23 (2016) 1365–1374 1369

Table 1

Amount of caffeine, catechins and amino acids in tea leaves that were sprayed with hot-water shower.

Treatment time (s) Caffeine (mg/g)

0 21.0 ± 0.42

180 ∗4.99 ± 0.22

280 ∗4.03 ± 0.12

Treatment time (s) Catechins (mg/g)

EGCG EGC ECG EC CG ( + )C Total

0 42 .6 24 .9 8 .57 8 .09 0 .19 0 .96 85 .3

± 0 .57 ± 0 .64 ± 0 .06 ± 0 .47 ± 0 .02 ± 0 .07 ± 1 .77

180 37 .1 22 .9 7 .47 6 .67 0 .15 0 .84 76 .1

± 0 .64 ± 0 .57 ± 0 .18 ± 0 .03 ± 0 .04 ± 0 .07 ± 1 .70

280 47 .8 26 .8 8 .9 8 .36 0 .11 0 .96 92 .9

± 4 .95 ± 3 .04 ± 0 .78 ± 0 .87 ± 0 .01 ± 0 .03 ± 9 .55

Time (s) Free amino acids (mg/g)

Theanine Glu Arg Asp Gln Ser Ala Asn GABA Total

0 7 .91 1 .53 3 .96 0 .80 1 .28 0 .42 0 .19 0 .06 0 .07 16 .2

180 7 .36 2 .34 4 .71 2 .04 1 .21 0 .62 0 .28 0 .31 0 .19 19 .1

280 6 .58 1 .97 4 .58 1 .17 0 .84 0 .72 0 .34 0 .08 0 .23 16 .5

The amount of each tea component in dried tea leaves was measured by HPLC.

( ∗p < 0.05)

EGCG, ( −)-epigallocatechin gallate; EGC, ( −)-epigallocatechin; ECG, ( −)-epicatechin gallate; EC, ( −)-epicatechin; CG,

( −)-catechin gallate; ( + )C, ( + )-catechin; Glu, glutamic acid; Arg, arginine; Asp, aspartic acid; Gln, glutamine; Ser,

serine; Ala, alanine; Asn, asparagine; GABA, γ -butyric acid.

Table 2

Amount of caffeine, catechins and amino acids in the eluate from each green tea.

Green tea Caffeine (mg/l)

Low-caffeine 0 .07

Theanine-rich 12 .1

Standard 5 .3

Green tea Catechins (mg/l)

EGCG EGC ECG EC CG ( + )C Total

Low-caffeine 0 .2 8 .0 0 .1 7 .3 0 0 .6 16 .1

Theanine-rich 0 .3 5 .3 7 .2 10 .4 0 .5 1 .9 25 .5

Standard 0 .2 7 .6 8 .1 14 .6 0 1 .3 31 .7

Green tea Free amino acids (mg/l)

Theanine Glu Arg Asp Gln Ser Ala Asn GABA Total

Low-caffeine 50 12 .0 13 .4 8 .0 10 .7 3 .2 1 .6 0 .7 1 .1 101

Theanine-rich 102 5 .8 8 .8 3 .6 5 .2 3 .5 1 .3 0 .8 0 131

Standard 18 4 .4 6 .6 2 .9 3 .9 1 .3 0 .5 0 .2 0 .4 39

Tea leaves (3 g) from each green tea were added to 1 l of water at room temperature, stirred for 6 min, then filtered.

The amount of each tea component in the eluate was measured by HPLC.

A

s

t

m

h

o

(

w

h

s

m

T

s

c

w

h

g

t

i

i

t

A

c

(

s

t

c

nti-stress effects of green tea on a mouse model of psychosocial

tress

The anti-stress effects of drinking low-caffeine Sencha,

heanine-rich Gyokuro, and standard Sencha were tested on a

ouse model of psychosocial stress evoked by confrontational

ousing. Adrenal hypertrophy, a typical stress response in living

rganisms, was used as a marker for assessing the degree of stress

Unno et al., 2011, 2013a ).

Each type of green tea was administered to mice as drinking

ater for 8 days. The adrenal weights of group and confrontational

ousing mice that consumed H 2 O or each green tea were mea-

ured ( Fig. 3 a). The changes of adrenal weight from group-housing

ice to confrontational-housing ones were compared ( Fig. 3 b).

he adrenal hypertrophy was significantly lower in mice that con-

umed low-caffeine and theanine-rich green teas than those that

onsumed water ( F (3, 20) = 9.560; p = 4.0 × 10 −4 ; η2 = 0.35; one-

ay ANOVA). To examine why significant suppression of adrenal

ypertrophy was not observed in mice of the standard green tea

roup, ingestion of each tea component was compared among

he three green tea groups of mice under confrontational hous-

ng. The drinking volume of the theanine-rich group was signif-

cantly higher than that of the standard and low-caffeine green

ea groups ( F (2, 15) = 17.661; p = 1.1 × 10 −4 ; η2 = 0.49; one-way

NOVA; Table 3 ). The ingestion of caffeine, a diuretic, was signifi-

antly higher in the theanine-rich green tea than in other groups

F (2, 15) = 573.58; p = 6.8 × 10 −15 ; η2 = 0.98; one-way ANOVA),

uggesting an increase in the volume of theanine-rich green tea

hat had been drunk. The ingestion of EGCG was also signifi-

antly higher in the theanine-rich group than in the other groups

1370 K. Unno et al. / Phytomedicine 23 (2016) 1365–1374

0

2

4

6

8

Water Low-caffeine Theanine-rich Standard

)g

m(t

hgie

wlaner

dA

Green tea

GroupConfrontation

(a)

0.0

0.5

1.0

1.5

2.0

2.5

Ch

ang

e o

f ad

ren

al w

eig

ht

(mg

)

Green tea

**

(b)

Fig. 3. Effects of consumption of green tea on adrenal gland weight. The adrenal weights of group housing (white column) and confrontational housing (gray column) mice

were measured (a). The changes of adrenal weight from group-housing mice to confrontational-housing ones were compared (b). Each bar shows the mean ± SD ( n = 6; ∗ p < 0.05).

K. Unno et al. / Phytomedicine 23 (2016) 1365–1374 1371

Table 3

Ingestion of green tea and main tea components consumed by mice.

Green tea Low-caffeine Theanine-rich Standard

Body weight (g) 32.58 ± 2.46 33.96 ± 1.19 33.62 ± 2.45

Ingestion volume (ml/day/mouse) 9.40 ± 0.95 ∗12.86 ± 1.20 9.40 ± 1.31

Caffeine (mg/kg) ∗0.02 ± 0.002 ∗4.6 ± 0.4 1.5 ± 0.2

EGCG (mg/kg) 0.06 ± 0.005 ∗0.11 ± 0.009 0.06 ± 0.007

EGC (mg/kg) 2.3 ± 0.5 2.0 ± 0.2 2.1 ± 0.5

Theanine (mg/kg) ∗14.5 ± 1.3 ∗38.6 ± 3.1 5.0 ± 0.6

Arg (mg/kg) ∗3.9 ± 0.4 * 3 .3 ± 0.3 1.8 ± 0.2

The value of each component of low-caffeine and theanine-rich green teas was compared

with that of the standard green tea group (mean ± SD, n = 6, ∗p < 0.05).

(

T

c

g

F

3

t

A

p

a

v

a

M

a

s

i

3

3

c

w

m

2

d

i

s

n

(

p

t

f

A

i

p

s

t

i

p

f

w

D

A

a

i

t

t

g

n

(

t

o

(

E

t

c

T

a

b

s

r

E

p

a

(

t

E

a

b

a

2

s

i

a

i

s

s

d

c

G

f

A

S

a

t

e

p

b

A

r

a

F (2, 15) = 121.74; p = 5.3 × 10 −10 ; η2 = 0.89; one-way ANOVA).

he ingestion of theanine and Arg was significantly higher in low-

affeine and theanine-rich groups than in the standard green tea

roup (theanine, F (2, 15) = 454.93; p = 3.8 × 10 −14 ; η2 = 0.97; Arg,

(2, 15) = 80.50; p = 9.5 × 10 −9 ; η2 = 0.84; one-way ANOVA, Table

). The effects of these tea components were then examined in de-

ail.

nti-stress effects of tea components on a mouse model of

sychosocial stress

Individual and combined effects of each tea component on

drenal hypertrophy were examined in 34 groups of mice. The

olumes that had been consumed were not significantly different

mong these groups. Mean volume was 10.00 ± 1.21 ml/mouse/day.

ean body weight, which was 31.28 ± 1.07 g, was not different

mong these groups.

First, the relationship between theanine and caffeine was as-

essed. Adrenal hypertrophy in mice under confrontational hous-

ng was significantly suppressed by drinking theanine (0.32 and

.2 mg/kg/day), but not in mice that ingested caffeine (0.32 and

.2 mg/kg/day) ( Fig. 4 a). While adrenal hypertrophy was signifi-

antly suppressed in mice that ingested theanine (3.2 mg/kg/day)

ith caffeine (0.32 mg/kg/day), the suppression was abolished in

ice that ingested theanine with high doses of caffeine ( F (8,

6) = 6.567; p = 1.1 × 10 −4 ; η2 = 0.45; one-way ANOVA). The dose-

ependent suppression of caffeine against theanine suggests the

mportance of a lower amount of caffeine in green tea for the anti-

tress effect of theanine.

EGC, which is 10 (w/w) times higher than theanine, did

ot suppress the effect of theanine, and low dose of EGC

3.2 mg/kg/day) showed an anti-stress effect ( F (6, 20) = 6.853;

= 4.6 × 10 −4 ; η2 = 0.45; one-way ANOVA; Fig. 4 b). Unlike EGC,

he gallate catechins EGCG and ECG strongly abolished the ef-

ect of theanine ( F (6, 20) = 1.219; p = 0.34; η2 = 0.07; one-way

NOVA; Fig. 4 c). Adrenal hypertrophy was significantly suppressed

n mice that ingested Arg but not Glu and Gln ( F (5, 20) = 11.972;

= 1.9 × 10 −5 ; η2 = 0.56; one-way ANOVA; Fig. 4 d).

The combined effects of theanine, caffeine, EGC, EGCG and Arg

howed that Arg significantly enhanced the anti-stress effect of

heanine in the presence of caffeine or EGCG on psychosocial stress

nduced adrenal hypertrophy in mice. On the other hand, EGC sup-

ressed the effect of EGCG, resulting in an enhanced anti-stress ef-

ect of theanine ( F (10, 33) = 4.433; p = 5.5 × 10 −4 ; η2 = 0.33; one-

ay ANOVA; Fig. 4 e).

iscussion

nti-stress effect of theanine, and counteracting effects of caffeine

nd catechins on theanine

Adrenal hypertrophy, a typical stress response in living organ-

sms, was significantly suppressed in stressed mice that ingested

heanine (0.32 mg/kg). This was much lower than the concentra-

ion of theanine in standard Sencha green tea (5.0 mg/kg), sug-

esting that other tea components suppress the effect of thea-

ine. Although caffeine has been reported to antagonize theanine

Kakuda et al., 20 0 0 ), we found that EGCG remarkably suppressed

he anti-stress effect of theanine. Whereas the oral bioavailability

f catechins and the distribution of EGCG in the brain is very low

Nakagawa et al., 1997; Suganuma et al., 1998 ), the suppression of

GCG against theanine suggests that EGCG was incorporated into

he brain and abolished the effect of theanine.

On the other hand, EGC had an anti-stress effect at a low con-

entration (3.2 mg/kg), and suppressed the effect of EGCG ( Fig. 4 e).

he relationship among theanine, EGCG and EGC suggests that EGC

ntagonizes EGCG, whereas EGCG antagonizes theanine. Since the

lood-brain permeability of EGCG was significantly lowered by the

ame molar amount of EGC using a model of the blood-brain bar-

ier (unpublished data), intestinal and blood-brain permeability of

GCG may be competitively suppressed by EGC. A lowered incor-

oration of EGCG into the brain results in lowered suppression

gainst theanine. However, EGC was needed at more than three

w/w) times higher level (or 5 times higher molar ratio) than EGCG

o compete inhibition, suggesting that EGC is a weak inhibitor of

GCG.

If so, then why did EGC have no effect on theanine? Catechins

re excreted by efflux transporters expressed on the cell mem-

rane ( Kadowaki et al, 2008 ) and the activity of efflux transporters

re suppressed by the presence of a gallate moiety ( Annaba et al.,

010; Farabegoli 2010 ). Efflux of EGC from the brain may reduce

uppression against theanine.

Whereas green tea is usually steeped in hot water and EGCG

s infused as the main catechin, non-gallate catechins such as EGC

nd EC are mainly eluted into water at room temperature. EGCG

s abundant in tea leaves; therefore, regulation of catechin compo-

ition by steep temperature is important for determining the anti-

tress effect of green tea.

The amount of caffeine was higher in Gyokuro than in stan-

ard Sencha. However, the relative amount of caffeine and gallate

atechin against theanine was higher in standard Sencha than in

yokuro. These results suggest that a balance among theanine, caf-

eine and catechins is important for anti-stress effect of green tea.

reduction of caffeine resulted in a relative increase of theanine in

encha. In addition, low-caffeine Sencha is useful for people such

s infants, pregnant women and elderly people who do not want to

ake a high dose of caffeine. If drinking green tea has similar ben-

fits for humans as mice, it is expected to reduce stress in many

eople by switching to the daily ingestion of green tea from other

everages.

nti-stress effects of amino acids in green tea

Theanine incorporated into brain is reported to reduce the

elease of glutamate from pre-synapse to the synaptic cleft by

cting as a Gln transporter and inhibiting the incorporation of

1372 K. Unno et al. / Phytomedicine 23 (2016) 1365–1374

Fig. 4. Individual and combined effects of theanine, other amino acids, caffeine and catechins on adrenal hypertrophy. Mice drank each concentration of theanine with

caffeine (a), theanine with EGC (b), theanine with gallate catechins such as EGCG and ECG (c), Arg, Glu and Gln (d), theanine with caffeine, EGCG, EGC and Arg (e), for 8 days

(single housing in a partitioned cage for 1 week followed by confrontational housing for 1 day). The weight of adrenal glands was compared. Each bar shows the mean ± SD

( n = 4; ∗ and #, p < 0.05). (a–d) Asterisk ( ∗) indicates statistical significance compared with control mice that ingested water under confrontational housing. e: # indicates

statistical significance compared with mice that ingested theanine (3.2 mg/kg), and ∗ indicates statistical significance between the two groups.

K. Unno et al. / Phytomedicine 23 (2016) 1365–1374 1373

e

G

c

h

w

s

2

G

t

a

o

t

w

p

c

e

c

i

p

m

t

n

t

2

(

t

s

n

w

M

t

a

t

t

1

g

C

t

p

c

A

T

i

a

c

t

t

h

a

c

t

d

C

e

c

o

A

e

f

t

S

f

R

A

F

G

G

H

I

I

K

K

K

K

K

K

M

M

M

M

M

M

N

S

S

xtracellular Gln into neurons, which suppresses the conversion of

ln to Glu by glutaminase ( Kakuda et al., 2008; Kakuda 2011 ). Glu

an be decarboxylated into γ -amino butyric acid (GABA). In the

ippocampus of mice that ingested theanine (6 mg/kg) in drinking

ater for 2 weeks, the levels of Glu and pyroglutamic acid were

ignificantly reduced, and conversely GABA increased ( Inoue et al.,

016 ), indicating that theanine modulates GABA production from

lu. Glu is the main excitatory neurotransmitter while GABA is

he main inhibitory neurotransmitter in the brain. Changes in Glu

nd GABA metabolism may play important roles in the control

f neuronal excitability. Since GABA in tea and foods is not able

o cross the blood-brain barrier and as such must be synthesized

ithin neurons ( Martin and Rimvall,1993 ), modulation of GABA

roduction by theanine would be important in the brain.

Whereas suitable synaptic excitation is important, excessive ex-

itation damages nerve cells and triggers neurodegenerative dis-

ases ( Mehta et al., 2013 ), suggesting that the suppression of ex-

essive Glu is important. On the other hand, excessive synaptic

nhibition may be attributed to Down syndrome, and EGCG sup-

resses over-expression of the GABA pathway in Down syndrome

ouse models ( Souchet et al., 2015 ). This data supports the no-

ion that theanine and EGCG modulate GABA production antago-

istically.

Arg is considered to be an important regulator in the cen-

ral nervous system through nitric oxide synthesis ( Virarkar et al.,

013 ) and has vital functions in physiological stress and anxiety

Gulati and Ray, 2014 ). Although the amount of Arg is lower than

hat of theanine ( Tables 1 and 2 ), Arg has a rather higher anti-

tress effect than theanine ( Fig. 4 ). High contents of both thea-

ine and Arg are important for the anti-stress effects of green tea,

hich is characteristic of high-grade green teas ( Mukai et al., 1992;

iyauchi et al., 2014 ). These results indicate that high-grade green

ea has an anti-stress effect. Theanine is the major amino acid in

ll tea cultivars. However, the difference in Arg content among cul-

ivars was larger than the other amino acids, and cultivars native

o Japan had more Arg than Assam hybrid cultivars ( Ikeda et al.,

993 ). A quality assurance study that considers the functions of

reen tea is necessary.

onclusions

Ingestion of green tea, such as low-caffeine Sencha and

heanine-rich Gyokuro that were eluted with water at room tem-

erature, exhibited an anti-stress effect on mice under stressful

onditions. In these infusions, the amount of theanine, EGC and

rg were relatively enriched and caffeine and EGCG were lowered.

he data of individual and combined effects of each tea component

ndicates that the ingestion of theanine and Arg, the main amino

cids in green tea, suppressed adrenal hypertrophy in mice, while

affeine and gallate catechins such as EGCG and ECG abolished

he anti-stress effects of theanine. In addition, non-gallate EGC re-

ained the anti-stress effect of theanine by suppressing EGCG.

Since Gyokuro exhibited anti-stress effects in mice, ingestion of

igh-grade green tea such as Gyokuro is highly expected to have

n anti-stress effect in humans. However, because the content of

affeine is high in Gyokuro, low-caffeine Sencha is a valuable green

ea to suppress stress in people who do not want to take a high

ose of caffeine.

onflict of interest

We wish to confirm that there are no known conflicts of inter-

st associated with this publication and there has been no signifi-

ant financial support for this work that could have influenced its

utcome.

cknowledgments

This research study was supported by a Grant-in-Aid for Sci-

ntific Research ( KAKENHI 23617014 and 15K00828 ) and a grant

or specially promoted research of the University of Shizuoka. We

hank Dr. Hara at University of Chicago for her valuable discussion.

upplementary materials

Supplementary material associated with this article can be

ound, in the online version, at doi:10.1016/j.phymed.2016.07.006 .

eferences

nnaba, F., Kumar, P., Dudeja, A.K., Saksena, S., Gill, R.K., Alrefai, W.A., 2010. Green

tea catechin EGCG inhibits ileal apical sodium bile acid transporter ASBT. Am.J. Physiol. Gastrointest. Liver Physiol. 298, G467–G473. doi: 10.1152/ajpgi.00360.

2009 .

arabegoli, F., Papi, A., Bartolini, G., Ostan, R., Orlandi, M., 2010. ( −)-Epigallocatechin-3-gallate downregulates Pg-P and BCRP in a tamoxifen

resistant MCF-7 cell line. Phytomedicine 17, 356–362. doi: 10.1016/j.phymed.2010.01.001 .

oto, T. , Horie, H. , Mukai, T. , 1993. Analysis of major amino acids in green tea byhigh-performance liquid chromatography coupled with OPA precolumn deriva-

tization. Tea Res. J. (Chagyo Kenkyu Hokoku) 77, 29–33 .

ulati, K., Ray, A., 2014. Differential neuromodulatory role of NO in anxiety andseizures: an experimental study. Nitric Oxide 43, 55–61. doi: 10.1016/j.niox.2014.

08.008 . orie, H. , Maeda-Yamamoto, M. , Ujihara, T. , Kohata, K. , 2002. Extraction of tea cate-

chins for chemical analysis. Tea Res. J. (Chagyo Kenkyu Hokoku) 94, 60–64 . keda, N. , Horie, H. , Mukai, T. , Goto, T. , 1993. Content of individual free amino acids

in the first and autumn flushes of cultivars for several tea types. Tea Res. J.(Chagyo Kenkyu Hokoku) 78, 67–75 .

noue, K., Miyazaki, Y., Unno, K., Min, J.Z., Todoroki, K., Toyo’oka, T., 2016. Stable

isotope dilution HILIC-MS/MS method for accurate quantification of glutamicacid, glutamine, pyroglutamic acid, GABA and theanine in mouse brain tissues.

Biomed. Chromatogr 30, 55–61. doi: 10.1002/bmc.3502 . adowaki, M., Sugihara, N., Tagashira, T., Terao, K., Furuno, K., 2008. Presence or ab-

sence of a gallate moiety on catechins affects their cellular transport. J. Pharm.Pharmacol. 60, 1189–1195. doi: 10.1211/jpp.60.9.0011 .

akuda, T., 2011. Neuroprotective effects of theanine and its preventive effects on

cognitive dysfunction. Pharmacol. Res. 64, 162–168. doi: 10.1016/j.phrs.2011.03.010 .

akuda, T., Hinoi, E., Abe, A., Nozawa, A., Ogura, M., Yoneda, Y., 2008. Theanine,an ingredient of green tea, inhibits [ 3 H] glutamine transport in neurons and

astroglia in rat brain. J. Neurosci. Res. 86, 1846–1856. doi: 10.1002/jnr.21637 . akuda, T. , Nozawa, A. , Unno, T. , Okamura, N. , Okai, O. , 20 0 0. Inhibiting effects of

theanine on caffeine stimulation evaluated by EEG in the rat. Biosci. Biotechnol.

Biochem. 64, 287–293 . imura, K. , Ozeki, M. , Juneja, L.R. , Ohira, H. , 2007. L-Theanine reduces psychological

and physiological stress responses. Biol. Psychol. 74, 39–45 . onishi, S. , Ozasa, M. , Takahashi, E. , 1972. Metabolic conversion of N-methyl car-

bon of γ -glutamylmethylamide to caffeine in tea plant. Plant Cell Physiol 13,365–375 .

aeda-Yamamoto, M. , Nagaya, H. , Mitsumori, T. , Yamaguchi, Y. , Horie, H. , Ema, K. ,

Suzuki, M. , Yamaguchi, H. , Warashina, F. , Mizukami, Y. , Hirono, H. , Monobe, M. ,2007. A change of chemical components and effect on anti-allergic activity in

‘Benifuuki’ green tea which was produced with low caffeine processing ma-chine. Jpn. J. Food Eng. 8, 109–116 .

artin, D.L. , Rimvall, K. , 1993. Regulation of gamma-aminobutyric acid synthesis inthe brain. J. Neurochem. 60, 395–407 .

ehta, A., Prabhakar, M., Kumar, P., Deshmukh, R., Sharma, P.L., 2013. Excitotoxicity:

bridge to various triggers in neurodegenerative disorders. Eur. J. Pharmacol. 698,6–18. doi: 10.1016/j.ejphar.2012.10.032 , 5 .

iyauchi, S., Yuki, T., Fuji, H., Kojima, K., Yonetani, T., Tomio, A., Bamba, T.,Fukusaki, E., 2014. High-quality green tea leaf production by artificial cultiva-

tion under growth chamber conditions considering amino acids profile. J. Biosci.Bioeng. 118, 710–715. doi: 10.1016/j.jbiosc.2014.05.008 .

onobe, M. , Ema, K. , Tokuda, Y. , Maeda-Yamamoto, M. , 2010. Effect on the epigallo-

catechin gallate/epigallocatechin ratio in a green tea ( Camellia sinensis L.) extractof different extraction temperatures and its effect on IgA production in mice.

Biosci. Biotechnol. Biochem 74, 2501–2503 . ukai, T. , Horie, H. , Goto, T. , 1992. Differences in free amino acids and total nitrogen

contents among various prices of green tea. Tea Res. J. (Chagyo Kenkyu Hokoku)76, 45–50 .

akagawa, K. , Okuda, S. , Miyazawa, T. , 1997. Dose-dependent incorporation of teacatechins, ( −)-epigallocatechin-3-gallate and ( −)-epigallocatechin, into human

plasma. Biosci. Biotechnol. Biochem. 61, 1981–1985 .

asaoka, K. , Kito, M. , Onishi, Y. , 1964. Synthesis of theanine by pigeon liver acetonepowder extract. Agric. Biol. Chem. 28, 325–330 .

himotoku, T. , Ichikawa, H. , Anan, T. , Takayanagi, H. , Ikegaya, K. , 1982. Relation be-tween amounts of some ingredients extracted from green tea and brewing con-

ditions. Tea Res. J. (Chagyo Kenkyu Hokoku) 55, 43–50 .

1374 K. Unno et al. / Phytomedicine 23 (2016) 1365–1374

U

U

V

Souchet, B., Guedj, F., Penke-Verdier, Z., Daubigney, F., Duchon, A., Herault, Y., Bi-zot, J.C., Janel, N., Créau, N., Delatour, B., Delabar, J.M., 2015. Pharmacological

correction of excitation/inhibition imbalance in Down syndrome mouse models.Front. Behav. Neurosci. 9, 267. doi: 10.3389/fnbeh.2015.00267 .

Suganuma, M. , Okabe, S. , Oniyama, M. , Tada, Y. , Ito, H. , Fujiki, H. , 1998. Wide distri-bution of [3H]( −)-epigallocatechin gallate, a cancer preventive tea polyphenol,

in mouse tissue. Carcinogenesis 19, 1771–1776 . Unno, K., Fujitani, K., Takamori, N., Takabayashi, F., Maeda, K., Miyazaki, H.,

Tanida, N., Iguchi, K., Shimoi, K., Hoshino, M., 2011. Theanine intake improves

the shortened lifespan, cognitive dysfunction and behavioural depression thatare induced by chronic psychosocial stress in mice. Free Radic. Res. 45, 966–

974. doi: 10.3109/10715762.2011.566869 .

nno, K., Iguchi, K., Tanida, N., Fujitani, K., Takamori, N., Yamamoto, H., Ishii, N.,Nagano, H., Nagashima, T., Hara, A., Shimoi, K., Hoshino, M., 2013a. Ingestion

of theanine, an amino acid in tea, suppresses psychosocial stress in mice. Exp.Physiol. 98, 290–303. doi: 10.1113/expphysiol.2012.065532 .

nno, K., Tanida, N., Ishii, N., Yamamoto, H., Iguchi, K., Hoshino, M., Takeda, A.,Ozawa, H., Ohkubo, T., Juneja, L.R., Yamada, H., 2013b. Anti-stress effect of thea-

nine on students during pharmacy practice: positive correlation among salivaryα-amylase activity, trait anxiety and subjective stress. Pharmacol. Biochem. Be-

hav. 111, 128–135. doi: 10.1016/j.pbb.2013.09.004 .

irarkar, M., Alappat, L., Bradford, P.G., Awad, A.B., 2013. L-arginine and nitric oxidein CNS function and neurodegenerative diseases. Crit. Rev. Food Sci. Nutr. 53,

1157–1167. doi: 10.1080/10408398.2011.573885 .