Differential effects of foods traditionally regarded as ‘heating’ and ‘cooling’ on prostaglandin E2 production by a macrophage cell line

(1)

~nal of

Biomedical

Science

Original Paper

J Biomed Sci 2002;9:596-606 DOl: 10.1159/000067288 Received: March 14, 2002 Accepted: May 6, 2002

Differential Effects of Foods Traditionally

Regarded as "Heating" and "Cooling" on

Prostaglandin E2 Production by a Macrophage

Cell Line

Ching-jang Huang Mei-Chiao Wu

Laboratory of Nutritional Biochemistry, Department of Agricultural Chemistry, National Taiwan University, Taipei, Taiwan, ROC

Key Words

COX-2 • Food • Macrophage • PGE2. Prostaglandin

production. Traditional food belief

Abstract

Some components of natural foods may enhance or

inhibit prostaglandin formation and potentially affect

the inflammation condition. A macrophage cell line,

RAW264.7, was employed to examine the effects of

foods traditionally regarded as 'heating' or 'cooling' on

the production of PGE2, a well-known proinflammatory

mediator. Foods traditionally regarded as 'heating' (lit-

chi, Iongan, and dried Iongan) or 'cooling' (chrysanthe-

mum flower, bitter gourd, and lotus seed plumule) were

extracted sequentially with water and ethyl acetate. The

water extracts (WE) and ethyl acetate extracts (EAE) were

applied to RAW264.7 macrophages in the presence or

absence of LPS (lipopolysaccharide). In the absence of

LPS, the WEs from the 'heating foods', litchi, tongan, or

dried Iongan had a dose-dependent enhancing effect on

PGE2 production, with respective EC50s of 8.4, 16, and

11 mg/ml. This effect was accompanied by significant

induction of COX-2 protein expression, as shown by

Western blot analysis. In contrast, LPS-induced PGE2

production was inhibited in a dose-dependent manner

by the WEs of the 'cooling foods', chrysanthemum flow-

er, bitter gourd, and lotus seed plumule, with respective

tC50s of 0.6, 0.13, and 0.08 mg/ml. At the concentrations

tested, none of the EAEs had any effect on basal PGE2

production, while LPS-induced PGE2 production was in-

hibited or increased by the EAE from bitter gourd and

Iongan, respectively. Water-soluble extracts of foods tra-

ditionally regarded as 'heating' enhanced basal PGE2

production, while those from 'cooling' foods significant-

ly inhibited LPS-induced PGE2 production by the macro-

phage celt line. This subject merits further study to deter-

mine whether appropriate food selection may help pa-

tients suffering from chronic inflammatory conditions.

Introduction

Prostaglandins (PGs), a family of intercellular and

intracellutar messengers derived from 20-carbon polyun-

saturated fatty acids, are synthesized in a wide range of

tissue types and serve as autocrine or paracrine mediators

signaling changes in the immediate environment. In hu-

mans, PGs are involved in diverse functions, including

blood clotting, ovulation, initiation of labor, bone metab-

olism, nerve growth and development, wound healing,

KAKG E R

Fax+41 61 306 12 34 E-Mail karger@karger.ch www.karger.com

1021-7770/02/0096-0596518.50/0 Accessible online at:

ww~'. karger.com/jbs

Ching-jang Huang, PhD

Department of Agricultural Chemistry National Taiwan University

1, Sec. 4, Roosevelt Road, Taipei, 106 Taiwan (ROC) TeL/Fax +886 2 23621301 E-Mail cjjhuang@ccmsmtu.edu.tw

(2)

kidney function, blood vessel tone, and the immune

response [42]. The key regulatory enzyme of P G biosyn-

thesis is cyclooxygenase (COX, or prostaglandin H syn-

thase, PGHS), a bifunctional enzyme catalyzing the oxi-

dative cyclisation of the central 5 carbons in 20-carbon

polyunsaturated fatty acids and subsequent peroxidation

[27, 55]. The final PGs produced vary, depending on the

downstream enzymatic machinery present in a particular

cell type.

Nonsteroidal anti-inflammatory drugs (NSAIDs) had

been used in humans for many decades before Vane [53]

proposed that the inflammation suppression produced by

NSAIDs might be due to inhibition of COX, resulting in

less proinflammatory P G production at the injury site.

Molecular cloning studies have identified two forms of

COX [15], COX-l, a constitutive form responsible for

'housekeeping' functions, and COX-2, an inducible form

that mediates many of the inflammatory and inducible

effects. Both forms carry out essentially the same catalytic

reaction and have similar primary sequences [15, 48].

Inhibitors acting only on the inducible COX-2 would

therefore be ideal as NSAIDs free of side effects [11].

PGE2 is the major P G produced by macrophages.

COX-2 expression in macrophages is induced in response

to intrinsic factors, such as cytokines, or extrinsic factors,

such as lipopolysaccharide (LPS), leading to the produc-

tion of PGE2 [45, 18]. PGE2 produced by macrophages is

a well-known pro-inflammatory mediator [ 17].

Some components of natural foods may enhance or

inhibit PG formation and potentially affect the inflamma-

tion condition. For example, phytochemicals and other

components in foods may modulate COX-2 expression

and P G formation [1, 5, 9, 20, 29, 32, 35, 50, 54]. It is

therefore worth screening the effect of foods on COX-2

expression and P G formation which can provide useful

information on food selection for the amelioration of the

pathophysiological conditions of inflammation.

For 2,000 years, people in Chinese and Indian societies

have believed that certain foods are either 'heating' (or

'fire increasing') or 'cooling' (or 'fire reducing') in the

body when eaten [2, 7, 14]. In addition, some physiologi-

cal conditions or disease states are categorized as 'hot' or

'cold'; but this is not related to body temperature. Accord-

ing to traditional Chinese medicine, symptoms for the

diagnosis of a 'heating' disease state include a dry mouth,

thirst, a bitter taste sensation even when not eating or

drinking, flushing, constipation, a small amount of dark-

colored urine, a rough yellow tongue, and a fast pulse [34].

Fever, overstimulation, inflammation, and sore throat

have also been described as conditions with excess heat in

the body [2]. In addition, acne or pimples, hemorrhoids,

and nosebleeds are also thought to be associated with an

excess consumption of heating food [22]. In contrast, dis-

ease states in which the patient is thirsty but is not willing

to drink, and has pallor, diarrhea, a large amount of

diluted urine, a smooth white tongue and a slow pulse are

diagnosed as 'cooling' [34]. Weakness, tiredness, cold

body temperature and shivering are also reported to be

'cooling' conditions [2, 14]. In agreement with the Chi-

nese beliefofyin and yang, health is thought to result from

a proper balance of 'heating' and 'cooling' foods and

activities, and illness is treated with foods or medicines

with properties opposite to those of the disease [2, t 4, 33].

In Taiwan, Hong Kong and among Chinese immigrants in

the USA, in which ahnost all of the population accept

m o d e m medical care, the concept of 'heating' and 'cool-

ing' foods is still prevalent in all socioeconomic classes.

Many individuals select food to avoid suffering from

unfavorable and disturbing conditions, often discounted

by most modern medical doctors [22, 33]. This is because

these traditional rules filled the explanatory and behavior

niches left open in Western medicine [22].

Not all foods are categorized as 'heating' or 'cooling'; a

large variety of foods such as staple foods like plain rice,

noodles and steamed bun are neutral [2, 7, 33]. Moreover,

foods can change categories as a result of different cooking

methods [2, 7, 33]. For example, masted and fried pea-

nuts are typical 'heating' foods, but boiled and steamed

peanuts are not. Fried foods and meat cooked in black

sesame oil with ginger are also good examples of 'heating'

foods produced by a cooking method. Most 'cooling'

foods are of plant origin, especially fruits and vegetables.

Many, but not all, of the 'cooling' fruits and vegetables

have a high water content, e.g. watermelon and radish,

and excretion of a large volume of diluted urine is consid-

ered one of the 'cooling' conditions. However, not all veg-

etables are 'cooling'; many are neutral, and a few fruits,

such as litchi, longan, and durian, are considered to be

'heating'.

In spite that food items listed as 'cooling' or 'heating'

may vary among societies or populations in different geo-

graphical areas because of different climates, environ-

ments and agricultural products available, the central rule

of the hot/cold system prevails. In addition, susceptibility

also varies among individuals which is partially attributed

to an innate 'hot' or 'cold' physical nature. Although there

has been speculation on the significance and rationale of

the 'hot/cold' food belief from a social science aspect [2-4,

14], an empirical rational basis has not been reported. A

chemical feature c o m m o n to 'hot' or 'cold' foods cannot

Heating/Cooling Foods and Macrophage

PGE2 Production

(3)

be ruled out simply based on the nutrient composition,

implying that some components with special functions

may be involved. On the other hand, the class of body

regulator that is targeted by 'hot' or 'cold' foods and

mediates the diverse 'hot/cold' syndrome must be one

that has a wide range of functions or can trigger diverse

physiological changes in the body.

Since (1) the extreme case of a 'heating' condition

resembles inflammation in certain ways, and (2) feeding

mice a diet containing frying oil tended to increase PGE2

production by peritoneal macrophages [26] and fried

foods are very common 'heating' foods [2, 33], the effect

of some typical 'heating' and 'cooling' foods on PGE2 pro-

duction was tested using a murine macrophage cell line,

RAW 2 64.7, as an in vitro model [38]. The 'heating'

tbods, litchi, longan, and dried longan [2], and the 'cool-

ing' foods, bitter gourd [14], chrysanthemum flower [2],

and lotus seed plumule, were chosen for testing because

these are natural foods in which the classification of 'heat-

ing' or 'cooling' is most widely accepted. Results of this

study will provide basic information supporting further

welt-controlled in vivo experiments.

Materials and M e t h o d s

Materials

The 'heating' foods chosen for this study were longan

(Nephelium

longana

Camb or

Euphoria longana

Lain), litchi (or lychee, leechee or lichi,

Litchi sinesis

Sonn), and dried longan (during drying, longan undergoes a marked browning reaction), while the 'cooling' foods chosen were dried chrysanthemum flower

(Chrysanthemum morifot-

ium

Hemsl.), fresh bitter gourd

(Momordica charantia

L.), and dried plumule of lotus seed

(Nelumbo nucifera

Gaertn). Longan, litchi, and bitter gourd, which are common fruit and vegetables in Taiwan, were purchased from a local market. Dried chrysanthemum flower and lotus seed plumule were purchased from a Chinese traditional medicine store. Dried longan, the hot-air-dried product of the whole longan fruit including the peel, was purchased from a grocery store.

The RAW264.7 macrophage cell line (CCRC60001, originally from the American Type Culture Collection; designation, TIB-71) was obtained from the cell bank of the Food Industry Research and Development Institute, Hsin Chu, Taiwan. Dulbecco's minimal essential medium (DMEM) and fetal bovine serum were purchased from Gibco (Md., USA).

Methods

Extraction of FoodSamptes.

The edible portions oflongan, litchi, dried longan (the aril), and bitter gourd (the flesh) were cut into smalI pieces and homogenized in a Waring blender with a minimal amount of double-distilled water, and the homogenates were filtered through several layers of gauze. Dried chrysanthemum flower was boiled in water (10 ml/g flower) for 5 rain aud filtered. Then the residue was reextracted in the same way and the two filtrates were combined. The dried lotus seed plumule was homogenized with water (10 ml/g plu-

mule) and filtered. All filtrates were centrifuged at 10,500 g at 4°C for 30 rain and the clear supernatant, the water extract (WE), was freeze-dried. The ethyl acetate extract (EAE) was prepared by freeze- drying the precipitate and filtered residues from the water extraction and extracting it overnight by shaking at room temperature with ethyl acetate (30 ml/g of residue), followed by filtration and drying of the filtrate in a rotary evaporator (Biichi, Essen, Germany). The yields of dried water extract of longan, litchi, dried longan, bitter gourd, lotus seed plumule and chrysanthemum were 7.15, 14.02, 43.49, 2.76, 28.95 and 34.5 g/100 g of the fresh edible portion (longan, litchi and bitter gourd) or dried material (dried longan, lotus seed plumule and chrysanthemum). The yields of evaporated EAE of longan, litchi, dried longan, bitter gourd and lotus seed plumule were 1.53, 4.01, 2.23, 1.86, and 17.25 g/100 g of the dried residue after water extraction.

For cell culture experiments, the WEs were dissolved in DMEM, while the EAEs were dissolved in a minimal amount of ethanol or dimethyl sulfoxide. Both were diluted to the appropriate concentration with DMEM immediately before use.

Cell Culture

RAW 264.7 cells were grown in 25-cm 2 flasks in DMEM containing 10% fetal bovine serum at 37 °C in a 5% CO2 atmosphere. Near- confluent cells were removed with a cell scraper, seeded on 96-well plates at a concentration of 6 x t0 s cell/ml, incubated for 24 h at 37 °C in 5% CO2, and then washed with phosphate-buffered saline (27 raM, pH 7.4, PBS). To test the effect of food extracts on PGE2 production, the cells were incubated for 18 h at 37 ° C with serum-free DMEM containing various concentrations of sample in the presence or absence of 1 ng/ml of LPS. Then the medium was collected and stored at -70 °C until analysis fbr PGE2. Cell viability was assayed using the MTT method [39]. Experiments were repeated at least three times in triplicate.

To study the effect of the above-mentioned treatment on COX-2 protein expression, confluent cells were seeded on 6-cm 2 dishes and treated for 12 h at 37°C, in the presence or absence of 1 ng/ml of LPS, with WEs or EAEs at the concentration that showed a maximal effect on PGE2 production and did not significantly affect cell viability. The cells were then washed twice with PBS, suspended in 1 ml of PBS using a cell scraper, and transferred to a microcentrifuge tube. After centrifugation, the supernatants were discarded and the cell pellets stored at -70 °C until analyzed by Western blotting.

Analyses

The collected medium was diluted to an appropriate concentration and analyzed for PGE2 by competitive enzyme-immunoassay using a commercial kit (Cayman, Ann Arbor, Mich., USA). For COX-2 protein analysis, cell pellets were lysed and the cellular pro- teins extracted, separated by polyacrylamide gel electrophoresis, and transferred to a polyvinylidene fluoride membrane, which was then treated sequentially with primary antibody (rabbit anti-COX-2 anti- serum, Cayman), secondary antibody (alkaline-phosphatase-conju- gated anti-rabbit antibody, Tropix, Applied Biosystems, Foster City, Calif., USA), and CDP-star (chemiluminescent substrate of alkaline phosphatase, Tropix). The specific protein band was visualized using X-ray film and quantitated by image analysis.

Statistical AnaIysis

The data for PGE2 production were expressed as the mean + SD for triplicate wells in a representative experiment. The significance of

(4)

differences at each sample concentration was analyzed by ANOVA and Duncan's multiple range test using SAS 5.1 software. To calculate the percentage enhancement of basal PGE2 production by a sample at a specific concentration, the maximal PGE2 production produced by this sample was taken as 100% and the enhancing effect

calculated using the equation: percent enhancement = [(PGE2 sample -

PGE2 basal)/(PGE2 maximum- PGE2 b~sal)] X 100%, where PGE2 basal is the PGE2 produced by nonstimulated untreated macrophages and PGE2 sample is the PGE2 produced in the presence of a specific concentration of the sample. The ECso value, calculated from the curve of

percentage enhancement versus concentration, is the concentration of sample that produces 50% enhancement. To calculate the percentage inhibition of LPS-stimulated PGE2 production, PGE2 production in the presence of 1 ng/ml of LPS was taken as 0% inhibition and basal PGE2 production as 100 % inhibition; the percentage inhibition by a sample at a specific concentration was calculated as: percentage

inhibition = [1 - (PGE2 s a m p l e + L P S - PGE2 basaI)/(PGE2 LPs -

P G E 2 basal)] X 1000/0. The ICs0 value, calculated from the curve of

percentage inhibition versus concentration, is the concentration of sample that results in 50% inhibition.

121

6 12

==9

g

u73

¢D 13- 12

g

~ 7 3 0 A F0 0.01 0.1 0.2 0.5 1 2 Concentration (mg/ml) b ab ab ab FO 0.01 0,05 0.1 Concentration (rng/rnl) i= ECs0 = 11 mg/ml 12 E o "c} O C}. u~ 3 - ( 9 r~ c F0 b c ~ c 1 2 5 10 Concentration (mg/ml) 1 2 ¸- 9 - £

g

"o o u~ 3 - (.9 e . B a a a a F0 0.02 0.05 0.1 0,2 Concentration (mg/ml) D ECs0 = 16 mg/ml 0.5 a

N

c c a d 0.2 F0 1 5 10 20 50 Concentration (mg/ml) F ECs0 = 8.4 mg/ml 12 9 v o r~

~ 3

(.9 IX. 20 F0 1 5 7.5 10 20 Concentration (mg/ml)

F i g . 1. Effect of WEs on basal PGE2 production. A Chrysanthemum flower. B Bitter gourd. C Lotus seed plumule. D Longan. E Dried longan. F Litchi. RAW264.7 cells (6 × 104 cells/well) were seeded on 96-well plates and grown for 24 h, then incubated for 18 h with DMEM or DMEM containing various concentrations of WE. PGE2 in the medium was then measured by competitive enzyme-immuno-

assay. The data are the mean + SD for triplicate wells from a representative experiment. The experiment was repeated at least three times with similar results. F0 indicates basal PGE2 production when cells were incubated in DMEM alone. Means not sharing a common letter (a-d) were significantly different (p < 0.05) when analyzed by ANOVA and Duncan's multiple range test,

Heating/Cooling Foods and Macrophage PGE2 Production

(5)

30 l

~25

2o4

E 0 0 ,.,? L9 n A ICso = 0.6 mg/ml 0 0.01 0.1 0.2 0.5 1 2 Concentration (mg/ml) C IC50 = 0.08 mg/ml a a a b 0 5e-3 0 . 0 1 0.05 0.1 0,2 Concentration (mg/mt) 0 0.5 1 5 10 20 Concentration (mg/ml) F0 b F0 F0

3° t

t

~ 1 0 L~ ~. 5

3o t

1

15 o K10 u7 5 30-

~

25 - - ~ 2 0 - O "~ 1 5 - O ~ . 1 0 - ~9 ~. 5- B IC50 = 0.13 mg/ml 0 0 . 0 1 0.02 0.05 0.1 0.2 Concentration (mg/ml) D 0 0.1 1 5 10 20 Concentration (rng/rnl) F b - r b 0 0.1 0,5 1 Concentration (m /ml) 0.5 F0 50 F0 10 c F0

Fig.

2. Effect of WEs on LPS-induced PGE2 production. A Chrysan- themum flower. B Bitter gourd. C Lotus seed plumule. D Longan. E Dried longan. F Litchi. RAW264.7 cells (6 x 104 cells/well) were seeded on 96-well plates and grown tbr 24 h, then incubated for 18 h with DMEM containing 1 ng/ml of LPS with or without various concentrations of the WEs. The PGE2 concentration of the medium was

measured using an enzyme-immunoassay kit. The data shown are the mean + SD for triplicate wells of a representative experiment. The experiment was repeated at least three times with similar results. Means not sharing a common letter (a-t) were significantly different (p < 0.05) when analyzed by ANOVA and Duncan's multiple range test.

R e s u l t s

The basal level of PGE2 production by nonactivated

macrophages (without LPS or any other stimulation) was

low and in the range of 0.5-1 ng/ml. As expected, treat-

ment with 1 ng/ml of LPS significantly increased the

expression of COX-2 protein and PGE2 production.

In this study, a wide range of concentrations of sample

extracts was tested, but the data shown are restricted to

those for experiments in which the concentration of

(6)

extract did not significantly affect cell viability, as deter-

mined by the MTT assay. The enhancing and inhibitor)'

effects described below were therefore not secondary to

any change in cell number or viability.

Effects of Food Extracts on PGE2 Production by

RA W264. 7 Cells

Water Extracts. The effects of the different WEs on

basal PGE2 production were first tested. As shown in fig-

ure 1, basal PGE2 production was unaffected by WEs

from chrysanthemum flower (fig. 1A), bitter gourd

(fig. 1B), and lotus seed plumule (fig. 1C), but was in-

creased in a dose-dependent manner by WEs from longan

(fig. 1D), dried longan (fig. 1E), and litchi (fig. 1F). At a

concentration of 50 mg/ml, longan WE increased PGE2

production 8.5-fold compared with the basal level, while,

at 20 mg/ml, the WEs of dried longan and litchi resulted

in 5- and 7.5-fold increases, respectively, over the basal

levels. The EC50s for the WE of longan, dried longan, and

litchi were 16, 11, and 8.4 mg/ml, respectively.

The WEs were then tested for their effect on LPS-

induced PGE2 production. As shown in figure 2, WE from

chrysanthemum flower (fig. 2A), bitter gourd (fig. 2B),

and lotus seed plumule (fig. 2C) inhibited LPS-induced

PGE2 production in a dose-dependent manner. At a con-

centration of 2 mg/ml, the WE from chrysanthemum

flower completely inhibited PGE2 production, while the

WEs from bitter gourd (0.5 mg/ml) and lotus seed plu-

mule (0.2 mg/ml) produced 68 and 63% inhibition,

respectively. At a concentration equal to or greater than

1.0 (bitter gourd) or 0.5 mg/ml (lotus seed plumule), the

WEs significantly reduced cell viability. The ICs0s for the

WEs from chrysanthemum flower, bitter gourd, and lotus

seed plumule were 0.6, 0. t 3, and 0.08 mg/ml, respective-

ly. In contrast, the WEs from longan (fig. 2D), dried lon-

gan (fig. 2E), and litchi (fig. 2F) did not inhibit LPS-

induced PGE2 production, but instead tended to increase

it.

Ethyl Acetate Extracts. Due to their limited solubility,

the EAEs could only be tested in the gg/ml concentration

range. At these levels, none of the EAEs significantly

altered basal PGE2 production (data not shown). How-

ever, when tested for effects on LPS-induced PGE2 pro-

duction, the EAE from bitter gourd showed dose-depen-

dent inhibition (ICs0 = 21 gg/ml, fig. 3A), and, at concen-

trations of 50-100 gg/ml, caused almost complete inhibi-

tion, while the EAE from longan resulted in increased

PGE2 production (fig. 3C).

Effects of Food Extracts on COX-2 Protein Expression

by tL4 W264. 7 Cells

Levels of COX-2 protein in RAW 264.7 cells treated

with various samples in the presence or absence of LPS

were determined by Western blot analysis. As shown in

figure 4A, WEs from litchi, longan, dried longan, and

lotus seed plumule induced significant COX-2 protein

expression in nonstimulated macrophages (without LPS).

None of the WEs had any significant effect on the amount

of COX-2 protein induced by 1 ng/mt of LPS (data not

shown); however the EAEs from lotus seed plumule and

longan (100 gg/ml) showed a significant reduction

(fig. 4B). As no significant effect on PGE2 production was

seen when 1L~W264.7 cells were treated with EAE alone,

the effect of EAEs on basal COX-2 protein was not

checked.

Discussion

Among the six foods tested in this study, WEs from the

'heating' foods, longan, litchi, and dried tongan, signifi-

cantly enhanced basal PGE2 production by the macro-

phage cell line RAW264.7 and also tended to increase

LPS-induced PGE2 production. In contrast, WEs from

the 'cooling' foods, bitter gourd, chrysanthemum flower,

and lotus seed plumule, did not stimulate basal PGE2 pro-

duction and inhibited LPS-induced PGE2 production.

These results support our hypothesis that 'heating' or

'cooling' foods may, respectively, stimulate or inhibit

PGE2 production by macrophages. It is therefore worth

testing these effects further in an appropriate in vivo mod-

el and also testing more foods for these effects.

The yields of dried water extract of longan, litchi and

bitter gourd corresponded to a WE concentration of 71.5,

140 and 27.6 mg/g of the fresh edible portion. This is

higher than their respective EC50s of 8.4 (litchi) and 16

mg/ml (longan), and IC50 of 0.13 mg/ml (bitter gourd).

Similarly, the ECs0 of WE from dried longan (11 mg/ml)

and ICs0s of the chrysanthemum flower (0.6 mg/ml) and

lotus seed plumule (0.08 mg/ml) were lower than the WE

concentration in 1% (wt/vol) chrysanthemum or dried

lotus seed plumule beverages or a 30/0 (wt/vol) dried lon-

gan beverage. After eating, the presence of the active com-

ponent at a sufficient concentration, at least in the gas-

trointestinal tract, is therefore possible.

Since the water-soluble component(s) of longan, litchi,

and dried longan stimulated PGE2 production by the

macrophage cell line and also tended to increase the

expression of COX-2 protein (fig. 4 A), it is probable that

(7)

IC50 = 21 gg/ml a b ~ b c

18 l

c 12 ~ c- o 9 " o o o_ 3 B - - - ' d d e o 0 5 10 25 50 100 FO 0 1 5 10 20 200 FO

Concentration (lag/ml) Concentration (gg/ml)

1 8 C 18 D 15 15 -£=1 ~ 1 2 - ~ a a a "o "13 O o " ~ 6 ~7 ~7 (.9 0 5 10 20 50 100 FO 0 1 2.5 5 10 50 FO Concentration (gg/ml) Concentration (gg/ml) 0 1 5 10 25 Concentration (gg/m() 50 FO

Fig. 3. Effect of EAEs on LPS-induced PGE2 production. A Bitter gourd. B Lotus seed plumule. C Longan. D Dried longan. E Litchi. The EAEs were dissolved in a minimal amount of dimethylsulfoxide or ethanol, then diluted with DMEM. The experiment was perfbrmed as in figure 2. Means not sharing a common letter (a-d) were significantly different (p < 0.05) when analyzed by ANOVA and Duncan's multiple range test.

the enhancing effect on PGE2 production involved induc-

tion of COX-2 expression.

COX-2 expression is induced in macrophages in re-

sponse to intrinsic factors such as cytokines, or extrinsic

factors such as LPS, leading to the production of PGE2

[ 18, 45]. Most of these factors are molecules that can trig-

ger the signaling pathway by binding to receptor mole-

ctdes in the cell membrane. For example, LPS acts on

macrophages by binding to LPS binding protein and

mCD14, a receptor located on the macrophage surface

membrane [51]. This activates the signal transduction

pathway, which includes protein kinase C [12] and

(8)

1 2 3 4 5 6 7 1 2 3 4 5 6 7 8 A 3 5 q WE(-LPS) S, 3 ¢'4

>k2

o ~ L o .> 1 2 3 4 5 B -$ 1.5 EAE (+LPS) Q. x o 1.o 0.5 % -~ 0 6 7 "" 1 2 3 4 5 6 7 8

Fig. 4. Western blot analysis of COX-2 protein expression atter treatment with extracts. A Treatment with WE or LPS alone: lane 1 = F0 (basal level, no treatment); lane 2 = FOL1, LPS alone (1 ng/ml); lane 3 = bitter gourd WE alone (0.2 mg/ml); lane 4 = dried longan WE alone (5 mg/ml); lane 5 = litchi WE alone (20 mg/ml); lane 6 = lotus seed plumule WE alone (0.2 mg/ml); lane 7 = longan WE alone (20 rag/ ml). B Treatment with LPS (1 ng/ml) with or without EAE: lane 1 = F0 (basal level, no treatment); lane 2 = FOL1 (LPS alone); lane 3 = LPS plus dried longan EAE (t 0 gg/ml); lane 4 = LPS plus litchi EAE (20 gg/ml); lane 5 = LPS plus bitter gourd EAE (100 l,tg/ml); lane 6 = LPS plus lotus seed plumule EAE (200 gg/ml); lane 7 = LPS plus longan EAE (20 btg/ml); lane 8 = LPS plus longan EAE (100 btg/ml).

RAW 264.7 cells (4 x 106/ml) were seeded on a 6-cm 2 dish and grown tbr 24 h. Then the medium was changed to DMEM or LPS alone or DMEM containing EAEs and LPS. The cells were then incubated for 12 h and harvested for Western blot analysis of COX-2 protein expression. The upper panels show Western blots fi'om a representative experiment; the experiment was repeated three times with similar results. The lower panels show image analysis results; the data are the means + SD for three experiments. * p < 0.05 vs. F0 (cells with no treatment) and + p < 0.05 vs. FOL1 (cell treated with LPS only). In the lower panel of A, the p value of the difference between longan (lane 7) and F0 (Lane 1) was 0.059, i.e. nearly significant.

MAPK [18]. Expression of the COX-2 gene is induced

when these signals activate transcriptional factors, such as

NF-~cB [10, 18, 31]. Theoretically, molecules able to bind

to cell surface receptors, trigger the signaling pathway, or

activate transcription factors could directly or indirectly

induce COX-2 protein expression and result in PGE2 pro-

duction. Molecules that interfere with any point in the

above-mentioned mechanism of action of LPS may lead

to inhibition of PGE2 production [19]. Other potential

factors that could influence PGE2 production include

those affecting phospholipase A2 activity [57] or the fatty

acid composition of the cell membrane [32], both of

which may alter substrate availability, or those that may

directly affect the catalytic activity of COX-2 [6].

Among the dietary and food factors that may affect

PGE2 production, direct inhibition of COX activity has

been demonstrated for resveratrol [20, 35, 50], hydroxy-

stilbenes [47], and curcumine [1]. Enhancement of ex-

pression of COX-2 protein has also been demonstrated by

n-6 fatty acids in the mammary gland [5]. Inhibition of

COX-2 expression in macrophages, epithelial cells, the

RAW 264.7 macrophage cell line, and colorectal cancer

cells has been reported to be caused, respectively, by

genistein [10], resveratrol [50], apigenin [29], and antioxi-

dants [9]; the first three agents are bioflavonoids with

antioxidant activity [44] and it appears that the antioxi-

dant activity is closely related to the inhibitory effect on

COX [35, 37, 52]. Using the same testing system as that

used in this study, we have observed a significant inhibi-

tion on the LPS-induced PGE2 production by a number of

phytochemicals including curcumine, quercetin, apige-

nin, narigenin, resveratrol, and some tea polyphenols

[54].

(9)

For chrysanthemum flower, we only tested the WE,

since it is usually ingested as a drink prepared by brewing

in hot water. When cells were treated with chrysanthe-

m u m flower WE, although inhibition of PGE2 production

was seen in the presence (fig. 2A) of LPS, no significant

effect was seen on COX-2 protein expression (data not

shown). These results suggest that chrysanthemum flower

WE may directly interfere with COX-2 enzyme activity.

Compounds isolated from chrysanthemum flowers, in-

cluding sesquiterpenes and flavones, have been shown to

inhibit LPS-induced nitric oxide production by macro-

phages [56] and to be antioxidative and inhibit cancer cell

growth [4t]. Some flavones may be partially soluble in

water. It is not known if antioxidant flavonoids are

responsible for this inhibition by chrysanthemum flower

WE.

Numerous biologically active components have been

identified in bitter gourd, also known as bitter melon. For

example, seed extracts contain MAP-30, an anti-HIV and

anti-tumor protein [28], momorcharins [43] and momor-

dins [21], which inactivate ribosomes, and momordin Ic

[30] and oleanolic acid glycoside [36], both of which can

alter gastrointestinal transit time and blood glucose. An

extract of bitter gourd fruit has been shown to have hypo-

glycemic activity [46] and to alter the activities of phase I

and phase I! drug metabolizing enzymes in the rat liver

[25]. In this study, the extract was prepared exclusively

from the fleshy part with the seeds removed. Both the WE

and EAE from bitter gourd caused dose-dependent inhibi-

tion ofLPS-induced PGE2 production; again, since COX-

2 protein expression did not change significantly, inhibi-

tion of COX-2 enzyme activity may have been responsi-

ble.

Lotus seed plumule WE showed a marked inhibitory

effect on PGE2 production on a dried weight basis. How-

ever, its effect on PGE2 production and COX-2 protein

expression was complex, since it significantly inhibited

LPS-induced PGE2 production (fig. 2C), but did not af-

fect COX-2 protein expression (data not shown). On the

other hand, a faint COX-2 protein band was seen on the

Western blot of RAW264.7 cells treated with lotus seed

plumule WE in the absence of LPS (fig. 4 A) and this was

accompanied by a very slight increase in PGE2 production

(fig. 1C). The most intriguing result was the effect of the

EAE, which significantly reduced COX-2 protein expres-

sion (fig. 4B), but did not significantly affect PGE2 pro-

duction (fig. 3 B) by RAW264.7 cells treated with LPS.

These results suggest that multiple compounds in lotus

seed plumule may exert diverse, or even opposite, effects

on the RAW264.7 macrophage cell line. Despite this com-

plicated result, the inhibitory effect of lotus seed plumule

WE on LPS-induced PGE2 production is considered most

significant, since brewing in hot water to prepare a drink

is the most common method used for the preparation of

this food material.

Depending on the nature of the inhibitor and the sys-

tem used, COX inhibitors have been shown to reduce [ 12,

24, 29, 50], induce [40], increase [8], or have no effect on

[6] levels of COX-2 mRNA or protein. Dietary bioflavon-

oids, such as genistein [12], apigenin [29], and resveratrol

[50], inhibit COX-2 gene expression, probably by interfer-

ing with the signal transduction pathway that leads to

expression of COX-2 genes. In contrast, COX inhibitors,

such as indomethacin or NS398, can induce COX-2

mRNA expression [40] in a manner similar to peroxisome

proliferators, or increase COX-2 protein levels [8], proba-

bly by stabilization of the enzyme protein. PGE2 can de-

stabilize the COX-2 protein [40]. In the presence of LPS,

the inhibition of PGE2 production without a change in

COX-2 protein levels produced by extracts of bitter

gourd, chrysanthemum flower, and lotus seed plumule are

similar to the results seen with NSAID reported by Bar-

rios-Rodiles et al. [6].

The molecule(s) in litchi, longan, and dried longan

responsible for induction of COX-2 expression in the

macrophage remain to be elucidated. Because of their

water-soluble nature, plant lectins may be candidates.

Lectins from various plant sources have been shown to

bind to macrophage surface glycoproteins [16, 23] and

can result in activation of the macrophages [23]. Further

studies are needed to test this possibility.

Macrophages represent a system of widely dispersed

cells that are able to recognize and destroy invading

microorganisms and altered host components, such as

apoptotic cells. The macrophage plays an important role

in the regulation of the immune response [ 13]. Fixed mac-

rophages, such as alveolar macrophages, Kupffer cells,

histocytes, mesangial cells, and microglial cells, like free

macrophages, are monocyte-derived cells and may serve

additional functions in the tissues in which they reside,

respectively. For example, mesangial cells are involved in

the regulation of the glomerular filtration rate and renal

blood flow [49]. In addition, the final products of COX,

such as PGE2 and other prostaglandins, are involved in

diverse physiological functions and pathophysiological

conditions. The results of this study provide one of several

possible rationales for the effects of 'heating' and 'cooling'

foods. However, given the diverse aspects of physiological

conditions associated with 'heating' and 'cooling', other

possible mechanisms of action await further discovery.

(10)

For example, diuretic and antidiuretic actions may be

physiological effects of these so-called 'heating' and 'cool-

ing' foods.

In conclusion, WEs of foods traditionally regarded as

'heating', namely litchi, longan, and dried longan, in-

creased COX-2 protein levels, and basal PGE2 production

by the RAW264.7 macrophage cell line. In contrast, WEs

of foods traditionally regarded as 'cooling', namely chry-

santhemum flower, bitter gourd, and lotus seed plumule,

significantly decreased LPS-induced PGE2 production by

these cells. These results provide preliminary evidence for

a scientific basis for the ancient oriental belief in 'heating'

and 'cooling' foods. As PGE2 produced by macrophages is

an important pro-inflammatory agent, this subject merits

further study, as this could provide useful information on

food selection as a means of treatment for the pathophysi-

ological condition of inflammation.

Acknowledgments

This study was supported by a grant (DOH89-TD-1065) from the D e p a r t m e n t of Health, Taiwan. W e t h a n k Dr. M i n - H s i u n g Lee for advice on food sample extraction and Mr. Kuo-Wey Chen o f the Institute of Food Science a n d Technology, N a t i o n a l Taiwan Univer- sity, for assistance with freeze-dl2eing of samples.

References

1 Ammon HP, Safayhi H, Mack T, Sabieraj J. Mechanism of anti- inflammatory actions of curcumine and boswellic acids. J Ethnophar- maco138:113-119;1993.

2 Anderson EN Jr. 'Heating 'and 'cooling' foods in Hong Kong and Taiwan. Soc Sci Inf 19:237- 268;1980.

3 Anderson EN Jr. 'Heating and cooling' foods re-examined. Soc Sci Meal 23:755-773;1984. 4 Anderson EN Jr, Why is humoral medicine so

popular'? Soc Sci Med 25:331-337;1987. 5 Badawi AF, E1-Sohemy A, Stephen LL, Gho-

shaI AK, Archer MC. The effect of dietary n-3 and i1-6 polyunsaturated fatty acids on the expression ofcyclooxygenase 1 and 2 and levels of p2 lras in rat mammary glands. Carcinogen- esis 19:905-910; 1998.

6 Barrios-Rodiles M, Keller K, Belley A, Chadee K. Nonsteroidal anti-inflammatory drugs inhibit cyclooxygenase-2 enzyme activity but not mRNA expression in human macrophages. Biochem Biophys Res Commun 225:896-900; 1996.

7 Briggs GM, Calloway DH. Nutrition and Phys- ical Fitness, ed 11. Philadelphia, Saunders, I984, chap 19, pp 481-494.

8 CaUejas NA, Castrillo A, Bosca L, Martin-Sanz P. Inhibition ofprostaglandin synthesis up-reg- ulates cyclooxygenase-2 induced by lipopolysaccharide and peroxisomal proliferators. J PharmacoI Exp Ther 288:1235-1241;1998. 9 Chinery R, Beauchamp RD, Shyr Y, Kirkland

SC, Coffey, RJ, Morrow JD. Antioxidants reduce cyclooxygenase-2 expression, prostaglandin production, and proliferation in colorectal cancer cetts. Cancer Res 58:2323-2327;1998. 10 D'Acquisto FD, Iuvone T, Rombola L, Saute-

bin L, Rosa MD, Carnuccio R. Involvement of NF-KB in the regulation of cyclooxygenase-2 protein expression in LPS-stimulated J774 macrophages. FEBS Lett 418:175-178; 1997. 11 Dewitt DL, Meade EA, Smith WL. PGH syn-

thase isozyme selectivity: The potential for saf- er nonsteroidal anti-inflammatory drugs. Am J Med 95:40S-44S; 1993.

12 Glaser KB, Sung A, Bauer J, Weichman BM. Regulation of eicosanoid biosynthesis in the macrophage. Involvement of protein tyrosine phosphorylation and modulation by selective protein tyrosine kinase inhibitors. Biochem Pharmacol 45:711-721; 1993.

13 Gordon S. The role of the macrophage in immune regulation. Res ImmunoI 149:685-688; 1998.

14 Gould-Martin K. Hot cold poison and dirt: Chinese folk medical categories. Soc Sci Med 12:39-46;1978.

15 Herschman HR. Regulation of prostaglandin synthase-1 and prostaglandin synthase-2. Can- cer Metastasis Rev 13:241-256; 1994. 16 Hoenig S, Skutelsky E, Leibovici J, Barot R.

Changes in distribution of Iectin receptors in macrophages activated by Nocardia water-soluble molecules. Cell Mol Biol 39:843-848; 1993.

17 Humes JL, Bonney R J, Pelus, L, Dahlgren ME., Kuehl FLA, Davies P. Macrophages synthesize and release prostaglandins in response to inflammatory stimuli. Nature 269:149-I51; 1977.

18 Hwang D, Jang BC, Yu G, Bludreau M. Ex- pression of mitogen-inducible cyclooxygenase induced by lipopolysaccharide. Mediation through both mitogen-activated protein kinase and NF-~B signaling pathways in macrophages. Biochem Pharmaco154:8%96; 1997. 19 Hwang D, Fischer NH, Jang BC, Tak H, Kim

JK, Lee W. Inhibition of the expression of inducible cyclooxygenase and proinflammatory cytokines by sesquiterpine lactones in macrophages correlates with the inhibition of MAP kinases. Biochem Biophys Res Commun 226: 810-818;1996.

20 Johnson JL, Maddipati KR. Paradoxical effects of resveratrol on the two prostaglandin H synthases. Prostaglandins Other Lipid Mediat 56:131-143;1998.

21 Kimura Y, Minami Y, Tokuda T, Nakajima S, Takagi S, Funatsu G. Primary structures of N- linked oligosaccharides of momordin-a, a ribo- some-inactivating protein from Momordica

charantia seeds. Agric Biol Chem 55:2031-

2036;1991.

22 Koo LC. The use of food to treat and prevent disease in Chinese culture. Soc Sci Med 18: 757-766;I984.

23 Krugluger W, Lucas T, Koller M, Boltz-Nitu- lescu G, Froster O. Soybean agglutinin binds a

1 6 0 - k D a rat macrophage membrane glycopro- tein and enhances cell differentiation and activation. Immunol Lett 52:53-56;1996. 24 Ku KM, Sansores-Garcia L, Chen XM, Mati-

jevic-Aleksic, N, DuM, Wu KK. Suppression of inducible cyclooxygenase 2 gene transcription by aspirin and sodium salicylate. Proc Natl Acad Sci USA 96:5292-5297; 1999.

25 Kusamran WR, Ratanavila A, Tepsuwan A. Effects of neem flowers, Thai and Chinese bitter gourd fruits and sweet basil leaves on hepat- ic monooxygenases and glutathione S-transfer- ase activities, and in vitro metabolic activation of chemical carcinogens in rats. Food Chem Toxico1122:121-126;1998.

26 Lai CC, Lin BF, Lin KW. Effects of high dietary fat and deteriorated frying oil on prostaglandin E2 production in BALB/c mice. J Chin Agric Chem Soc 35:401-412; 1997. 27 Lands WEM. "Iqae biosynthesis and metabo-

lism of prostaglandins. Annu Rev Physiol 4l: 633-652;1979.

28 Lee-Huang S, Huang PL, Chen HC, Bourin- baiar A, Huang HI, Kuang HF. Anti-HIV and anti-tumor activities of recombinant MAP30

from bitter melon. Gene 161:151 - 156; 1 9 9 5 .

29 Liang YC, Huang YT, Tsai SH, Lin-Shiau SY, Chen CF, Lin JK. Suppression of inducible cyclooxygenase and inducible nitric oxide synthase by apigenin and related flavonoids in mouse macrophage. Carcinogenesis 20:1945- I952;1999.

Heating/Cooling Foods a n d Macrophage PGE2 P r o d u c t i o n

(11)

30 Li Y, Matsuda H, Yamahara J, Yoshikawa M. Acceleration of gastrointestinal transit by momordin Ic in mice: Possible involvement of 5- hydroxytryptamine, 5-HT(2) receptors and prostaglandins. Eur J Pharmacoi 392:71-77; 2000.

31 Lo CJ, Cryer HG, Fu M, Lo FR. Regulation of macrophage eicosanoid generation is dependent on nuclear factor ~B. J Trauma 45:19-24; 1998.

32 Lokesh BR, Kinsella JE. Modulation of prostaglandin synthesis in mouse peritoneal macrophages by enrichment oflipids with. either eico- sapentaenoic or docosahexaenoic acids in vitro. Immunobiology 175:406-419;1987. 33 Ludman EK, Newman JM. Yin and yaug in the

health-related food practices of three Chinese groups. J Nutr Educ 19:3-5;1984.

34 Ma CC. Chung I Chert Tuan Hseueh (Diagno- sis in Chinese Traditional Medicine). Taipei, Kuo Lee Bian I Kuan. 139;1980.

35 Martinez J, Moreno JJ. Effect ofresveratrol, a natural polyphenolic compound, on reactive oxygen species and prostaglandin production. Biochem Pharmaco159:865-870;2000. 36 Matsuda H, Li Y, Murakami T, Matsumura N,

Yamahara J, Yoshikawa M. Antidiabetic prin- ciples of natural medicines. IlL Structure-related inhibitory activity and action mode ofolea- nolic acid glycosides on hypoglycemic activity. Chem Pharm Bull 46:1399- t 403; 1998. 37 Mendez C, Garcia I, Maier, R. Antioxidants

attenuate endotoxin-induced activation of alveolar macrophages. Surgery 1 I8:412-420; 1995.

38 Miwa M, Kong ZL, Shinohara K, Watanabe M. Macrophage stimulating activity of foods. Ag- ric BioI Chem 54:1863-1866; 1990.

39 Mosmann T. Rapid colorimetric assay for cellular growth and survival: Application to proliferation and cytotoxicity assay. J Immunol Methods 65:55-63;1983.

40 Pang L, Hoult JR. Induction ofcyclooxygenase and nitric oxide synthase in endotoxin-activated J774 macrophages is differentially regu- lated by indomethacin: Enhanced cyclooxygenase-2 protein expression but reduction of inducible nitric oxide synthase. Eur J Pharmacol 317:151-155;1996.

41 Peterson J, Dwyer J. Flavonoids: Dietary oc- currence and biochemical activity. Nutr Res

18:1995-2018;1998.

42 Dubois RN, Abramson SB, Crofford L, Gupta RA, Simon LS, Van De Putte LBA, Lipsky PE. Cyclooxygenase in biology and disease. FASEB J 12:1063-1073;1998.

43 Pu Z, Lu BY, Liu WY, Jin SW. Characteriza- tion of enzymatic mechanism of gamma-mo- morcharin, a novel ribosome-inactivating protein with lower molecular weight of 11,500

purified from the seeds of bitter gourd (Mo-

mordica charantia). Biochem Biophys Res

Commun 229:287-294; 1996.

44 Rice-Evans CA, Miller NJ, Paganga G. Struc- ture-antioxidant activity relationships of flavonoids and phenolic acids. Free Radic Biol Med 20:933-956;1996.

45 Riese J, HoffT, Nordhoff A, Dewitt DL, Resch K, Kaever V. Transient expression of prostaglandin endoperoxide synthase-2 during mouse macrophage activation. J Leukocyte Biol 55: 476-482; 1994.

46 Sarkar S, Pranava M, Marita R. Demonstra-

tion of the hypoglycemic action of Momordica

charangia in a validated animal model of dia-

betes. Pharmacol Res 33:1-4; 1996.

47 Shin NH, Ryu SY, Lee H, Min KR, Kim Y. Inhibitory effects of hydroxystilbenes on cyclooxygenase from sheep seminal vesicles. Planta Med 64:283-284; 1998.

48 Smith WL, Garavito RM, DeWitt D L Prosta- gtandin endoperoxide H synthases (cyclooxygenase)-I and -2. J Biol Chem 271:33157- 33t60;1996.

49 Stock JD, Sansom SC. Regulation of filtration rate by glomerular mesangial cells in health and diabetic renal disease. Am J Kidney Dis 29: 971-981;1997.

50 SubbaramaiaI K, Quing W, Michaluart P, Te- lang N, Tanabe T, Inoue H, Jang M, Pezzuto JM, Dannenberg AJ. Resveratrot inhibits cyclooxygenase-2 transcription and activity in phorbol ester-treated human mammary epithelial cells. J Biol Chem 273:21875-21882;1998. 51 Sweet M J, Flume DA. Endotoxin signal trans-

duction in macrophages. J Leukocyte Biol 60: 8-26;1996.

52 Tetsuka T, Baler LD, Morrison AR. Antioxi- dants inhibit interleukin-1 induced cyclooxygenase and nitric oxide synthase expression in rat mesangial cells. Evidences for post-transcriptional regulation. J Biol Chem 271:11689- 11693;1996.

53 Vane JR. Inhibition ofprostaglandin synthesis as a mechanism of action for aspirin-like dlnags. Nature 231:232-235;1971.

54 Wu MC, Huang, CJ. Inhibition of prostaglandin E2 production of a macrophage cell line by some phytochemicals. Food Sci Agric Chem 3: 59-71;2001.

55 Yamamoto S. Characterization of enzymes in prostanoid synthesis. In: Curtis-Prior PB, ed. Prostaglandins: Biology and Chemistry of Pros- taglandins and Related Eicosanoids. Edin- burgh, Churchill Livingstone. 37-51; 1988. 56 Yoshinkawa M, Morikawa T, Toguchida I,

Harima S, Matsuda H. Medicinal flowers. 1I. Inhibitors of nitric oxide production and abso- lute stereostructures of five new germacrane- type sesquiterpenes, kikkanols D, D monoacetate, E, F, and F monoacetate from the flowers of Chrysanthemum indicum L. Chem Pharm

Bull 48:651-656; 2000.

57 Zor U, Reiss N. Regulation of plasma membrane phospholipase A2 activity by phosphorylation/dephosphorylation: Is glucocorticoid action mediated by induction of protein phosphatase. In: Bailey JM, ed. Prostaglandins, Leuko- trienes, Lipoxins and PAF. New York, Plenum Press. 135-139;1991.