permeability studies of kavalactones using a caco-2 cell monolayer model

ORIGINAL ARTICLE

Permeability studies of Kavalactones using a Caco-2 cellmonolayer model

A. Matthias* PhD, J. T. Blanchfield� PhD, K. G. Penman* PhD, K. M. Bone*� BSc (Hons)

Dip Phyt, I. Toth�§ PhD DSc and R. P. Lehmann*� PhD

*MediHerb Research Laboratories, Brisbane, Queensland, Australia, �School of Molecular and MicrobialSciences, The University of Queensland, Brisbane, Queensland, Australia, �School of Health, University ofNew England, Armidale, Australia and §School of Pharmacy, The University of Queensland, Brisbane,Queensland, Australia

SUMMARY

Objective: To examine the bioavailability of

kavalactones in vitro and the possible differences

in their bioavailability because of variations in

either chemical structure or the method of

extraction used.

Research design and methods: Caco-2 cell mono-

layers were used to determine the potential

bioavailability of kavalactones. Kavalactones

were added to the apical layer and basolateral

samples were taken over 150 min to examine

the concentration diffusing across the cell

monolayer. Kavalactone concentrations in these

samples were determined by high pressure liquid

chromatography.

Results: Kavalactones were found to be poten-

tially bioavailable as they all readily crossed the

Caco-2 monolayers with apparent permeabilities

(Papp) increasing from 42 · 10)6 cm/s and most

exhibiting more than 70% crossing within 90 min.

Not all differences in their bioavailability can be

related to kavalactone structural differences as it

appears that bioavailability may also be affected

by co-extracted compounds. For example, the Papp

for kawain from ethanol extracts was higher than

the values obtained for the same compound from

water extracts or for the kavalactone alone.

Conclusions: While the extraction method used

(ethanol or water) influences the total (but not the

relative) concentrations of kavalactones, it does

not markedly affect their bioavailability. Hence,

any differences between an ethanolic or an

aqueous extract in terms of the propensity of kava

to cause liver damage is not because of differing

kavalactone bioavailabilities.

Keywords: bioavailability, Caco-2 monolayers,

kava, kavalactone, Piper methysticum

INTRODUCTION

Kava is a traditional beverage and psychotropic

drug used by South Pacific Islanders for both

ceremonial and medical purposes. It is derived

from the rootstock of a sterile cultivated species of

Piper methysticum and contains natural products

from a number of phytochemical groups including

kavalactones and chalcones (flavokavains) as well

as small amounts of essential oil.

Traditionally, potent beverages are prepared by

chewing or pounding the root to produce a cloudy,

milky mash which is then consumed orally. The

major physiological action in humans is that of a

mild centrally acting relaxant (1, 2), with a numbing

of the mucous membranes of the mouth and tongue

noted when the beverage is consumed. In Europe, it

was popularly consumed as a treatment for anxiety

as clinical observations have revealed increased

calmness, a reduction of tension, irritability and

anxiety as well as improvements in concentration

and efficiency in intellectual and practical work in

patients consuming kava extracts (3). In a review of

six kava clinical trials (n = 345 subjects), meta-ana-

lysis revealed significant (P = 0Æ01) reductions in

anxiety, using the Hamilton anxiety scale, with

kava as compared with placebo (4).

Received 06 July 2006, Accepted 2 January 2007

Correspondence: Assoc. Prof. R. P. Lehmann, MediHerb

Research Laboratories, 3/85 Brandl Street, Eight Mile Plains,

Brisbane, Queensland 4113, Australia. Tel.: +61 7 3423 6521; fax:

+61 7 3423 6599; e-mail: [email protected]

Journal of Clinical Pharmacy and Therapeutics (2007) 32, 233–239

� 2007 The Authors. Journal compilation � 2007 Blackwell Publishing Ltd 233

The psychosedative property of kava has been

attributed to the kavalactones. They have been

shown to be responsible for the action of kava in a

number of animal models (5, 6). Kavalactones are a

group of structurally related lipophilic lactone

derivatives with an arylethylene-alpha-pyrone

skeleton (Fig. 1) and can represent from 3 to 20% of

the dried rootstock of the plant depending upon

the age of the plant and specific cultivar. Chemical

differences amongst these kavalactones are mainly

in the 5,6 and 7,8 double bonds as well as the

presence or absence of phenyl ring substituent

groups.

Despite the demonstrated benefits of kava, in

2002 its therapeutic use was banned in Germany,

Japan, France, Canada and the UK because of

reported cases linking kava consumption with

hepatotoxicity. All the reported cases were

for kava products where the extract had been

prepared using organic solvents such as ethanol

and acetone. It has been suggested that some

differences in the kavalactone profile that are

encountered when different extraction solvents

are used (7) may be responsible for these

hepatotoxic reactions.

One of the aims of this research was to under-

stand if an aqueous extract of kava might result in a

different absorption profile for the kavalactones

compared with an ethanolic extract. The effect of

different extraction procedures has thus been

compared in the event that other co-extracted

compounds (that may differ according to the

extraction solvent used) could affect the potential

bioavailability of the different kavalactones. We

have also examined variations in the bioavailability

of the kavalactones caused by differences in their

chemical structures. Alterations in bioavailability

because of extraction solvent, kavalactone profile

or chemical structure could provide an explanation

for the link between cases of liver damage and the

consumption of kava extracted with organic

solvents such as ethanol.

METHODS

Cell culture

Caco-2 cells were obtained from the American

Type Culture Collection (Rockville, MD, USA).

Transwell polycarbonate inserts were from Costar

(Cambridge, MA, USA) and cell culture reagents

were purchased from Gibco-BRL (Gran Island, NY,

USA).

Caco-2 cells were maintained in Dulbecco’s

modified Eagle’s medium (DMEM) at 95%

humidity and 37 �C in an atmosphere of 5% CO2,

(a)

R1

R2

O O

O

CH3

(b)

R1

R2

O O

O

CH3

(c)

R2

O O

O

CH3

R1

Fig. 1. Kavalactone structures. (a) dihydrokawain

R1 = R2 = H and dihydromethysticin R1 + R2 = OCH2O;

(b) kawain R1 = R2 = H and methysticin R1 + R2 =OCH2O; (c) desmethoxyyangonin R1 = R2 = H and

yangonin R1 = H, R2 = OCH3 and dehydromethysticin

R1 + R2 = OCH2O.

� 2007 The Authors. Journal compilation � 2007 Blackwell Publishing Ltd, Journal of Clinical Pharmacy and Therapeutics, 32, 233–239

234 A. Matthias et al.

supplemented with 10% foetal bovine serum,

2 mM LL-glutamine and 1% non-essential amino

acids. The medium was changed every second day.

After reaching 80% confluence, the cells were

subcultured using 0Æ2% EDTA and 0Æ25% trypsin.

Approximately 1 · 10)6 cells (passage 26) were

seeded onto polycarbonate cell culture inserts

(Transwell, mean pore size = 0Æ45 lm, 6Æ5 mm

diameter) and cultivated in the described supple-

mented DMEM also containing 100 units/mL

penicillin and 100 lg/mL streptomycin changed

every second day. The cells were allowed to grow

and differentiate for 21 days. Transepithelial elec-

trical resistance (TEER) of the monolayers was

measured using the Millicell-ERS system (Millipore

Corp., Bedford, MA, USA) before and after trans-

port experiments.

Transport experiments

Permeability assays were performed, as previously

described (8), in Hank’s balanced salt solution

(HBSS) containing 25 mM 4-(2- hydroxyethyl)-1-

piperazineethanesulfonic acid (HEPES) (pH 7Æ4) in

air at 95% humidity and 37 �C. Prior to the study,

the monolayers were washed in prewarmed HBSS–

HEPES for 30 min. At the start of the experiments,

100 lL of HBSS–HEPES containing the test pre-

paration was added to the apical and 600 lL of

HBSS–HEPES was added to the basolateral side

of the monolayers. The plates were shaken in a

Heidolph Titramax (Heidolph Instruments, Nurn-

berg, Germany) 1000 at 400 rpm at 37�C through-

out the experiment. At 10, 20, 30, 60, 90, 120 and

150 min, 400 lL of the basolateral volume was

removed and replaced with fresh HBSS–HEPES.

The apical solution was sampled only at the con-

clusion of the experiment.

Sample preparation

Two kava extracts were obtained from MediHerb,

Warwick Australia. Both extracts were prepared

from milled kava root/rhizome with one extracted

using a hot water protocol while the other was

extracted using 96% ethanol. Kawain was pur-

chased from Extrasynthese (France). Stock solu-

tions were made in 70% ethanol and filtered

(0Æ45 lm) prior to dilution in HEPES buffer and

addition to the apical side of the monolayers.

Analysis

Kavalactone concentrations in the samples were

determined by high pressure liquid chromatogra-

phy using a gradient system (Shimadzu LC10AT;

Shimadzu Corp., Tokyo, Japan). The mobile phase

was a mixture of 50 mMM phosphoric acid and

acetonitrile. The elution gradient used was 30–80%

acetonitrile over 30 min, followed by a period of re-

equilibration at 30% acetonitrile prior to the next

injection. An Alltima C18, 5 l, 150 · 4Æ6 mm

(Alltech; Deerfield, IL, USA) column was used with

a solvent flow rate of 0Æ8 mL/min.

Determination of permeability coefficients

The apparent permeability coefficients (Papp, cm/s)

were determined according to the following

equation:

Papp ¼ ðdC/dt� VrÞ=AC0

where dC/dt is the steady-state rate of change in the

compound concentration in the basolateral/receiver

chamber, Vr is the volume of the receiver chamber, A

is the surface area of the cell monolayers and C0 is the

initial concentration in the apical/donor chamber.

Four replicates of each test solution were performed.

RESULTS

Caco-2 monolayer integrity and the validity of the

data generated were assured by the measurement

of TEER data at the beginning and end of the

experiment. Monolayer TEER values were

260 ± 10 X cm2 at the start of the experiment and

280 ± 20 X cm2 when measured at the conclusion

of the experiment. This indicates that the mem-

branes were intact and not adversely affected by

the compounds.

The kava extracts used in these experiments

contained a range of kavalactones (methysticin,

dihydromethysticin, kawain, dihydrokawain,

dehydromethysticin, desmethoxyyangonin and

yangonin) as shown in Fig. 2. There were only

minor difference in the relative concentrations of

each kavalactone found in both the water and

ethanol extracts examined in this study. The com-

position of each apical preparation is given in

Table 1. The difference between the two extracts

was in the total amount of kavalactones per gram

of extract, with the ethanol extract containing a


Permeability studies of Kavalactones 235

total of 204 mg/g kavalactones and the water

extract containing a total of 103 mg/g kavalac-

tones.

Transport kinetics for the kavalactones are

shown in Fig. 3. The kavalactones readily diffused

across the Caco-2 monolayer although the

permeability varied, depending on the structure.

The ability of the kavalactones to cross the mono-

layers, the percentage transport at 90 min and the

calculated Papp values are given in Table 2. After

90 min, more than 70% of the kavalactones (except

yangonin) added to the apical layer diffused across

the Caco-2 cell monolayer. Yangonin was only 35%

diffused after 90 min. More than 82% of all

kavalactones, except yangonin, added to the apical

layer were able to be found in either the basolateral

fractions or remaining in the apical layer for both

extracts and the kawain standard. Yangonin was

only recovered at approximately 40% for both

extracts. The two major kavalactones found in both

extracts, kawain and methysticin, had higher Papp

values (>240 · 10)6 cm/s) which were three- to

sixfold greater than the Papp values calculated for

the other kavalactones.

Potential synergistic effects of the other kav-

alactones on the ability of kawain to cross the

monolayer was also examined and the results are

depicted in Fig. 4. In the absence of the other kav-

alactones, 84 ± 3% of the total kawain added to the

apical layer at a concentration of 22Æ9 lg/mL had

crossed the monolayer after 90 min. This is not

significantly different to the percentage that dif-

fused in the presence of other kavalactones for the

two extracts examined. The calculated Papp values,

however, do indicate an effect of the presence of

other kava components. The Papp of

302 ± 8 · 10)6 cm/s calculated for kawain alone

was significantly lower (P < 0Æ001) than the values

of 349 ± 7 and 387 ± 3 · 10)6 cm/s (Table 2) for

kawain in the presence of the other kavalactones in

either the ethanol or water extracts. The synergistic

effects also differ depending on the extraction

conditions used, as seen with the higher Papp value

obtained for the ethanol extract.

Table 1. Kavalactone apical concentrations

Compound

Apical

concentration

(lg/mL) % of Total

Water Ethanol Water Ethanol

(1) Methysticin 31Æ0 19Æ4 34 23

(2) Dihydromethysticin 4Æ4 4Æ1 5 5

(3) Kawain 27Æ2 30Æ0 30 35

(4) Dihydrokawain 6Æ4 8Æ9 7 10

(5) Dehydromethysticin 8Æ1 8Æ1 9 9

(6) Desmethoxyyangonin 6Æ8 8Æ5 8 10

(7) Yangonin 6Æ0 6Æ7 7 8

Total kavalactones 89Æ9 85Æ7

0

20

40

60

80

100

0 40 80 120 160

Time (min)

% A

pic

al

Fig. 3. Kavalactone transport kinetics in Caco-2 mono-

layers during a 2Æ5-h incubation. Methysticin (d), dihy-

dromethysticin ( ), kawain ( ), dihydrokawain (),

dehydromethysticin ( ), desmethoxyyangonin ( ),

yangonin ( ). Values are mean ± SD for n = 4. Error bars

not visible are within the symbol.

Minutes

1

2

3

4

5 67

0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34

mA

U0

2040

6080

100

120

140

160

180

200

Fig. 2. Kava HPLC trace at 220 nm of showing the

investigated kavalactones. 1 – methysticin, 2 – dihydro-

methysticin, 3 – kawain, 4 – dihydrokawain, 5 – dehy-

dromethysticin, 6 – desmethoxyyangonin, 7 – yangonin.



DISCUSSION

Kavalactones were found to be potentially

bioavailable as they all readily crossed the Caco-2

monolayers, most with more than 70% crossing

within 90 min. The apparent permeability (Papp)

values were calculated from the percentage uptake

data from 10 to 90 min with all Papp calculated to

be >40 · 10)6 cm/s. Papp values >1 · 10)6 cm/s

are considered to indicate almost complete intesti-

nal absorption (9, 10). This indicates that all the

kavalactones are able to cross the intestinal barrier.

Although all kavalactones appear to be bioavaila-

ble, there is still a six- to ninefold variation in the

values obtained, depending on the kavalactone

examined and whether it was from the water or

ethanol extract.

The Papp values obtained for the kavalactones in

this study are within the range (10–85 · 10)6

cm/s) obtained by Avdeef et al. (11) using filter

immobilized artificial membranes. In contrast to

their study however, where several kavalactones

were found to be markedly retained in the mem-

branes, only yangonin was potentially retained in

the Caco-2 monolayers. Only 40% of the yangonin

added to the apical layer was able to be recovered

from either the apical or basolateral volumes. All

other kavalactones were able to be recovered at a

minimum of 82%.

The two kava extracts examined in the present

study were of very similar relative kavalactone

composition (see Table 1), when calculated as a

percentage of the total kavalactones present. This is

unlike the water and ethanol extracts prepared and

compared by Cote et al. (7). However, similar to

their extracts (7), was the finding that the ethanol

extract examined in this study contained nearly

twice the amount of kavalactones per mg of extract

than did the water extract. As such, for the pur-

poses of this experiment, the water extract added to

the apical side of the Caco-2 cells was prepared

from twice the mass of extract as compared with

Table 2. Kavalactone permeability across Caco-2 monolayers after 90 min

Compound

% Uptake at 90 min

P value

Papp (·10)6 cm/s)

at 90 min

P valueWater Ethanol Water Ethanol

(1) Methysticin 74 ± 2 77 ± 1 NS 376 ± 9 241 ± 3 <0Æ001

(2) Dihydromethysticin 71 ± 1 76 ± 1 <0Æ001 52 ± 1 50 ± 1 0Æ01

(3) Kawain 80 ± 2 81 ± 1 NS 349 ± 7 387 ± 3 <0Æ001

(4) Dihydrokawain 83 ± 2 83 ± 2 NS 84 ± 3 116 ± 1 <0Æ001

(5) Dehydromethysticin 81 ± 3 95 ± 3 <0Æ001 116 ± 6 128 ± 3 0Æ01

(6) Desmethoxyyangonin 66 ± 12 70 ± 7 NS 85 ± 12 108 ± 6 0Æ01

(7) Yangonin 35 ± 3 32 ± 2 NS 42 ± 3 44 ± 2 NS

Data are mean ± SD (n = 4).

Differences between the Papp values for the water and ethanol

Kava preparations were calculated using the Student’s t test.

NS, not significantly different.

0

20

40

60

80

100

0 40 80 120 160

Time (min)

% A

pic

al

WaterEthanolStandard

*

*

Fig. 4. Effect of extraction method and the presence of

other kavalactones on Kawain permeability. Values are

mean ± SD for n = 4. Error bars not visible are within the

symbol. Error bars not visible are within the symbol.

* P < 0Æ05 when compared with the ethanol preparation

using the Student’s t test.



the ethanol extract, so that the same total concen-

tration of kavalactones was available to permeate

across the monolayer.

There are two characteristics that appear to affect

the permeability of the kavalactones. The first is the

presence of the methoxy (OCH3) group at R2 and

the absence of an oxygen functionality at R1 on

yangonin (see Fig. 1). It seems that the methoxy

group significantly decreases the permeability of

the kavalactone, decreasing Papp from

85 · 10)6 cm s for desmethoxyyangonin to

42 · 10)6 cm/s for yangonin in the water extract.

This is similar to the lower brain concentrations

found for these two kavalactones compared with

those for kawain and dihydrokawain after inter-

peritoneal injections at the same concentration (12).

The dioxole (OCH2O) group joining R1 and R2 as

well as the double bonds in the pyran ring and on

the link between the two ring structures have no

consistently significant effect on the permeability

differences between the kavalactones in this model.

The second factor that influences the permeability

of the kavalactones appears to be the other com-

ponents present in the extract. Purified kawain has

a Papp of 302 · 10)6 cm/s which is significantly

lower than that obtained for either of the extracts

examined and the kawain in the ethanol extract has

a significantly higher Papp than that found for

kawain in the water extract. All of this suggests

that the permeability of the kavalactones can be

influenced by other compounds co-ingested with

them. Those compounds in the extracts that influ-

ence kawain bioavailability are present in greater

amounts in the ethanol extract (assuming they are

the same compounds). Different compounds with

differing effects on kawain bioavailability could be

present in each extract (water or ethanol) and this

effect may be compound rather than concentration

driven.

This study has shown that although there are

potential bioavailability differences between the

different kavalactones, all should be able to cross

the intestinal barrier giving high plasma concen-

trations. It appears that the extraction method used

is able to influence the total concentrations of

kavalactones present in a preparation but does not

markedly effect the bioavailability of these kav-

alactones. Hence, any differences between an

aqueous and an ethanolic extract of kava, in terms

of a propensity to trigger idiosyncratic liver

damage, does not appear to be related to relative

kavalactone concentrations or bioavailability.

ACKNOWLEDGEMENTS

This work was funded by MediHerb Pty. Ltd.

REFERENCES

1. Cairney S, Maruff P, Clough AR, Collie A, Currie J,

Currie BJ (2003) Saccade and cognitive impairment

associated with kava intoxication. Human Psycho-

pharmacology: Clinical and Experimental, 18, 525–533.

2. Lehrl S (2004) In sleep disturbances associated with

anxiety disorders. Results of a multicentre, rand-

omized, placebo-controlled, double-blind clinical

trial. Journal of Affective Disorders, 1490 101–110.

3. Kretschmer W (1981) The effects of DD, LL-Kavain.

Agressologie, 22, 23.

4. Pittler MH, Ernst E (2003) Kava extract for treating

anxiety. Cochrane Database Systemic Review, 1,

CD003383.

5. Garrett KM, Basmadjian G, Khan IA, Schaneberg BT,

Seale TW (2003) Extracts of kava (Piper methysticum)

induce acute anxiolytic-like behavioural changes in

mice. Psychopharmacology, 170, 33–41.

6. Feltenstein MW, Lambdin LC, Ganzera M, Dha-

rmaratne HRW, Nanayakkara NPD, Khan IA, Sufka

KJ (2003) Anxiolytic properties of Piper methysticum

extract samples and fractions in the chick social-

separation-stress procedure. Phytotherapy Research,

17, 210–216.

7. Cote CS, Kor C, Cohen J, Auclair K (2004) Compo-

sition and biological activity of traditional and

commercial kava extracts. Biochemical and Biophysical

Research Communications, 322, 147–152.

8. Matthias A, Blanchfield JT, Penman KG, Toth I, Lang

C-S, De Voss JJ, Lehmann RP (2004) Permeability

studies of alkylamides and caffeic acid conjugates

from echinacea using a Caco-2 cell monolayer

system. Journal of Clinical Pharmacy and Therapeutics,

29, 7–13.

9. Artursson P, Karlsson J (1991) Correlation between

oral drug absorption and apparent drug permeabil-

ity coefficients in human intestinal epithelial

(Caco-2) cells. Biochemical and Biophysical Research

Communications, 175, 880–885.

10. Artursson P, Palm K, Luthman K (2001) Caco-2

monolayers in experimental and theoretical predic-

tions of drug transport. Advanced Drug Delivery

Reviews, 46, 27–43.

11. Avdeef A, Strafford M, Block E, Balogh MP, Cham-

bliss W, Khan I (2001) Drug absorption in vitro



model: filter-immobilized artificial membranes. 2.

Studies of the permeability properties of lactones in

Piper methysticum forst. European Journal of Pharma-

ceutical Sciences, 14, 271–280.

12. Keledjian J, Duffield PH, Jamieson DD, Lidgard RO,

Duffield AM (1988) Uptake into mouse brain of four

compounds present in the psychoactive beverage

kava. Journal of Pharmaceutical Sciences, 77, 1003–1006.



permeability studies of kavalactones using a caco-2 cell monolayer model

Documents