534 Notes Biol. Pharm. Bull. 31(3) 534—537 (2008) Vol. 31, No.

The Flavouring Phytochemical 2-Pentanone Reduces Prostaglandin

Production and COX-2 Expression in Colon Cancer Cells
Jenny PETTERSSON,a Pernilla Christina KARLSSON,b Ulf GÖRANSSON,a Joseph James RAFTER,b and
Lars BOHLIN*,a
a
Division of Pharmacognosy, Department of Medicinal Chemistry, Biomedical Centre, Uppsala University; Box 574, SE-
751 23 Uppsala, Sweden: and b Division of Medical Nutrition, Department of Biosciences and Nutrition, Karolinska
Institutet; Novum, SE-141 86 Stockholm, Sweden.
Received August 31, 2007; accepted December 4, 2007; published online December 6, 2007

Many phytochemicals found in the diet may prevent colon carcinogenesis by affecting biochemical processes
in the colonic mucosa. Inflammation and subsequent elevation of the enzyme cyclooxygenase-2 (COX-2) are two
such factors involved in the development of colon cancer, and inhibition of these processes could be important
targets for chemoprevention. We have previously shown COX-2 inhibitory activity locally in the colon; e.g. in
human fecal water from a group of vegetarians. In this study we focus on 2-pentanone, a frequently occurring
compound in common foods such as banana and carrot. The aim was to study the inhibitory effects on
prostaglandin production and COX-2 protein expression in tumour necrosis factor-a stimulated colon cancer
cells (HT29) by radioimmunoassay and Western blotting. 2-Pentanone inhibited both prostaglandin production
and COX-2 protein expression in human colon cancer cells. A concentration of 400 m mol/l 2-pentanone inhibited
the prostaglandin production by 56.9
12.9% which is in the same range as the reference compound NS398
(59.8
7.6%). The two highest concentrations of 2-pentanone were further analyzed by Western blot, and
400 m mol/l and 200 m mol/l 2-pentanone resulted in a 53.3
9.6% and
27.1% reduction of the COX-2 protein
levels respectively. Further studies on flavouring compounds, for example 2-pentanone, as colon cancer chemo-
preventives would be very valuable, and such results may contribute to future dietary recommendations.
Key words cyclooxygenase-2 inhibitor; prostaglandin E2; 2-pentanone; inflammation; colon cancer cell

Cyclooxygenases (COX) are enzymes involved in the pro- genic effects.20—23)

duction of prostaglandins and are hence of importance for A vegetarian diet, which is rich both in fibers and phyto-
many processes in the human body.1) COX-1 regulates nor- chemicals, is believed to protect against development of can-
mal body homeostasis whereas COX-2 is inducible and plays cer in the colon.15,24) We have previously studied naturally
a significant role in inflammatory processes. COX-2 is nor- occurring compounds and their inhibitory effects on COX-2,
mally sparsely expressed in the colon, but it can be induced and a variety of compounds have shown ability to affect the
by growth factors, cytokines, tumor necrosis factor (TNF) enzymatic activity of COX-2.17,20,25,26) Recently, we observed
and lipopolysaccharides in situations of stress.2—4) Elevated COX-2 inhibitory effects of human fecal water obtained from
levels of COX-2 enzyme in the colon are associated with vegetarians.27)
colon cancer.5,6) During the last decade, there has been a Now, we have extended our studies to flavouring com-
growing interest in nonsteroidal anti-inflammatory drugs pounds. One such compound is 2-pentanone, a small volatile
(NSAIDs) and COX-2 inhibitors as they clearly prevent de- molecule naturally occurring in fruits and vegetables, such as
velopment of cancer in the colon.7,8) The most convincing banana, cultivated strawberries and carrot root.28—31) 2-Pen-
protective effects have been demonstrated for selective COX- tanone has also been detected in soya oil, apple aroma con-
2 inhibitors.9—11) In 1999 the U.S. Food and Drug Adminis- centrate and in cheese.32—34) The amount of 2-pentanone
tration approved the use of celecoxib (a selective COX-2 in- (pentan-2-one) detected in different banana cultivars ranged
hibitor) for patients with familial adenomatous polyposis from 4.8 to 6.0 mg/kg.28) The average human daily consump-
(FAP), with the aim to prevent colon cancer development in tion of banana is 15.7 g, which corresponds to a daily intake
these patients.12) However, several recent studies have re- of 424 m g.35) The amount of 2-pentanone in carrot is about
vealed that selective COX-2 inhibitors can give rise to severe 200 m g/kg fresh carrot root.31) Moreover, 2-pentanone is pro-
side effects such as an increased risk of cardiovascular duced by the fungus Rhizopus oligosporus during the soy-
events.13) As an alternative to these synthetic selective com- bean tempeh fermentation process and by the common food
pounds, natural COX-2 inhibitors in the diet could be of im- contaminant fungi Penicillium verrucosum and Aspergillus
portance in cancer chemoprevention. niger.36—38) Staphylococcus carnosus, used as a starter in
Phytochemicals with anti-inflammatory and chemopreven- sausage fermentation, produces 2-pentanone and other methyl
tive properties are present in human food, such as vegetables, ketones, which contribute to the cured aroma of sausage
spices and fruit.14—16) Many of these have been investigated meat.39) 2-Pentanone has also been identified in the air phase
for their COX-2 inhibitory effects in vivo and in vitro.17) Cur- of human urine samples and as a volatile odorant from swine
cumin and resveratrol are examples of dietary compounds facilities.40,41) In addition, anaeorbic bacteria isolated from
that have been extensively studied for their anti-inflammatory fresh water sediment can produce 2-pentanone as a byprod-
and anticarcinogenic properties.18,19) Flavouring compounds, uct of sorbic acid degradation.42) Furthermore, 2-pentanone
such as cinnamaldehyde, diallylsulfides and eugenol have has been identified as a volatile fragment from cholesterol
also been studied for their COX-2 inhibiting and anticarcino- autoxidation and as one of many volatile compounds de-
∗ To whom correspondence should be addressed. e-mail: lars.bohlin@fkog.uu.se © 2008 Pharmaceutical Society of Japan
March 2008 535

tected in heated beef fat.43,44) Untreated cells and cells treated with only TNF-a were in-
Taken together there is major evidence that 2-pentanone is cluded in each experiment. After treatment, cell extracts were
frequently occurring in food of different kinds and hence we prepared and proteins were isolated. In brief, the cells were
were interested in investigating the potential anti-inflamma- washed with PBS, trypsinated and collected in tubes. The
tory effects of this compound. The aim with the present cells were concentrated by centrifugation and again washed
study was to investigate whether the flavouring compound 2- with PBS. After a second centrifugation step the cells were
pentanone exhibits any COX-2 inhibitory effects in vitro. lysed by addition lysis buffer (25 mM Tris acetate EDTA
(TAE), 10% glycerol, 1% Triton X-100, 1 mM EDTA and 2 mM
MATERIALS AND METHODS dithiothreitol (DTT), pH 7.8), followed by freezing (20 °C).
The protein concentration was determined using the Brad-
Materials 2-Pentanone was obtained from Acros, GTF ford protein assay and the protein concentration was adjusted
(Göteborg, Sweden). Arachidonic acid, aspirin, prostaglandin to 80 m g/sample.49) Proteins were separated by SDS-PAGE
E2, anti-prostaglandin E2, TNF-a , bovine serum albumin and (8%) as described by Laemmli50) and blotted onto a Hybond
charcoal were obtained from Sigma-Aldrich (Stockholm, enhanced chemiluminescence nitrocellulose membrane
Sweden). NS398 ([N-[2-(cyclo-hexyloxy)-4-nitrophenyl]- (Amersham). The membranes were incubated for 2 h with
methanesulfonamide]) was from Cayman Chemical Co. (Ann mouse monoclonal antibody against COX-2 (Transduction
Arbor, MI, U.S.A.). DMEM (Dulbecco’s Modified Eagle’s Laboratories, dilution 1 : 500) and actin (Santa Cruz Biotech-
Medium—high glucose) and Trypsin-EDTA were purchased nology, dilution 1 : 500). Thereafter the membranes were in-
from Invitrogen (Taastrup, Denmark). [3H]Prostaglandin E2 cubated with polyclonal anti-mouse IgG conjugated with
and Dextran T70 were obtained from Amersham Biosciences horseradish peroxidase (Dakopatts, dilution 1 : 2000). The
(Uppsala, Sweden). Stock solutions of arachidonic acid, as- ECL system was used to detect the bands, and emitted
pirin, prostaglandin and 2-pentanone respectively were pre- light was quantified using a FUJI LAS 100. Actin was used
pared by dissolving the compounds in EtOH. The final as an internal control for equal protein concentration loading
ethanol concentration in the cell media did not exceed 0.5%. per lane. Protein bands were quantified densiometrically and
Cell Culture The human colon carcinoma cell line expressed as the percentage inhibition of the TNF-a -treated
HT-29 (American type Culture Collection) was cultured cell band intensity (COX-2/actin). The values are means

in monolayer in Dulbecco’s Modified Eagle’s Medium S.E.M., n3. Values were normalized for enabling compari-
(DMEM) (supplemented with 10% fetal bovine serum son between different experiments. 2-Pentanone was tested in
(FBS), 2 mmol/l L-glutamine, and 1% penicillin/strepto- 2 different concentrations; 200 m mol/l and 400 m mol/l. These
mycin) at 37 °C and 5% CO2. Experiments were carried out concentrations were selected based on the first studies on
in supplemented medium (DMEM) containing 0.1% FBS. PGE2 production.
Prostaglandin E2 (PGE2) Production in HT29 Cells
PGE2 is a major product produced by COX from arachidonic RESULTS
acid and is often used to estimate COX activity in cells.45,46)
Cells were plated in 12-well plates and cultured to 60—80% 2-Pentanone inhibited prostaglandin production in a dose-
confluence. To reduce the activity of COX-1, cells were pre- dependent manner in TNF-a stimulated HT29 cells. The
treated with 100 m mol/l aspirin for 10 h. At day 3, the cells highest concentration (400 m mol/l) inhibited the 2-pentanone
were incubated with TNF-a (50 m g/l) and 2-pentanone (25— inhibited prostaglandin production by 56.9
12.9% (Fig. 1)
400 m mol/l) for 5 h. After treatment, the test solutions were which is almost the same inhibition as observed for the refer-
removed and replaced with 100 m mol/l arachidonic acid ence compound NS398 (59.8
7.6%). Also, 2-pentanone sig-
(Sigma) diluted in 0.1% DMEM and left for 1 h. The PGE2
concentration in the cell supernatants was determined by
radio immuno assay (RIA) using [3H]PGE2 and polyclonal
antiserum to PGE2 (Sigma). The samples were assayed undi-
luted and a standard curve with PGE2 was included in each
run. Each 2-pentanone concentration was tested at least twice
in the cell system and later analyzed in duplicate in the RIA.
The results were expressed as the percentage inhibition of the
TNF-a treated cells. The values are presented as means

S.D., n4. The values were normalized for enabling compar-
ison between different experiments. In all experiments un-
treated cells were included as controls, and the selective
COX-2 inhibitor NS398 was used as a reference compound
for comparison of inhibiting activity. The 2-pentanone con-
centrations tested were evaluated for cytotoxicity in the Ala-
marBlueTM assay to ensure that COX-2 inhibitory effects Fig. 1. 2-Pentanone Decreased Prostaglandin Production in TNF-a
Treated HT29 Cells
were not due to cell death.47,48)
Results are presented as relative inhibition of prostaglandin production in cells (val-
COX-2 Protein Production in HT29 Cells 5.2106 ues are normalized for an easy comparison between different experiments). HT29 cells
cells were seeded out in petri dishes. At 60—80% confluency stimulated with TNF-a (50 m g/l) for 5 h (TNF); cells simultaneously treated with 25—
400 m mol/l 2-pentanone and TNF-a (50 m g/l) (25—400); cells simultaneously treated
the cells were incubated with TNF-a (50 m g/l) in the pres- with NS398 (25 m mol/l) and TNF-a (50 m g/l) (NS398). Values are means
S.E.M.,
ence or absence of 2-pentanone (200—400 m mol/l) for 5 h. n3.
536 Vol. 31, No. 3

In a previous study we investigated the COX-2 inhibitory

effects of vegetarian fecal water and found that human fecal
water contains components that can affect both the COX-2
enzymatic activity and protein expression in HT29 cells.27)
Preliminary results indicate that the COX-2 inhibitory effect
is not caused by phenolic compounds (unpublished data). We
were therefore interested in identifying the compounds re-
sponsible for the anti-inflammatory activity of the fecal water
samples. The naturally occurring compound 2-pentanone is a
small molecule found in fruits and vegetables, and may be
one of the preventive anti-inflammatory factors.28—31) To our
knowledge there are no previous reports about cellular ac-
tions of 2-pentanone.
In the present study the in vitro COX-2 inhibitory effects
of 2-pentanone were investigated in TNF-a stimulated colon
cancer cells. Previous results show that fecal water samples
from vegetarians decreased both PGE2 production (5.4—
39.7% inhibition) and COX-2 protein levels (19—63% inhi-
bition).27) Compared to these results, 2-pentanone showed a
more pronounced COX-2 inhibitory effect at the concentra-
Fig. 2. 2-Pentanone Inhibited COX-2 Protein Production in TNF-a Stimu- tions tested. The results from the AlamarBlueTM assay
lated HT29 Cells showed that the cell viability was not extensively reduced
Untreated cells (1); cells treated with TNF-a (50 m g/l) (2); cells simultaneously after treatment with 2-pentanone. With a survival index of
treated with TNF-a (50 m g/l) and 2-pentanone (400 m mol/l) (3); cells simultaneously 84—96%, all 2-pentanone concentrations tested can be con-
treated with TNF-a (50 m g/l) and 2-pentanone (200 m mol/l) (4); All cells were incu-
bated with TNF-a and test component for 5 h. The Western blot is from a single experi- sidered to have a very low or negligible cytotoxic effect.51)
ment but is representative of at least 3 independent experiments. Protein bands were The level of inhibitory effect on PGE2 production is not
quantified densitometrically and expressed as a percentage of the untreated cell band
intensity (COX-2/actin) (upper panel). Values are means
S.E.M., n4. likely to be solely explained by the cytotoxic response, as the
COX-2 inhibitory effect of 2-pentanone remained on the pro-
tein level detected with Western blot carried out on cell ex-
tracts of living cells. Compounds (natural or synthetic) may
affect COX-2 at several levels, including inhibition of enzy-
matic activity, or by altering the levels of COX-2 protein or
mRNA. COX-2 inhibition can act through many different
mechanisms, and it is difficult to predict the exact mecha-
nism in this case. From our results we can conclude that the
PGE2 production among other factors may be partly ex-
plained by a cytotoxic effect. However it should be empha-
sized that COX-2 inhibition was clearly detected on the pro-
tein expression level.
Several sources in our diet seem to contain small amounts
of 2-pentanone.28—34) The total concentration in the colonic
Fig. 3. All 2-Pentanone Concentrations Tested Showed Low Cytotoxicity lumen is difficult to estimate, but it probably varies due to the
When Evaluated in the AlamarBlueTM Assay composition of the diet. The presence of 2-pentanone in the
The cell viability is presented as percentage of cell survival; survival index (SI). colonic lumen would also be dependent on the metabolic fate
of the compound during the gastrointestinal passage.
nificantly reduced the COX-2 protein production in the More comprehensive studies on the actual concentrations
human colon cancer cell line HT29. A concentration of in the lumen are needed to confirm the presence of 2-pen-
400 m mol/l 2-pentanone resulted in a 53.3
9.6% reduction tanone in fecal water. However, it cannot be excluded that
and 200 m mol/l in a 38.9
27.1% reduction of COX-2 protein fecal water contains other phytochemicals that contribute to
levels in TNF-a stimulated cells (Fig. 2). Cell viability was the observed COX-2 inhibitory effect. Small compounds,
tested to confirm that the effects were not due to cytotoxicity. such as flavouring agents, are widely present in our diet, and
The cell viability was only slightly reduced by treatment of their effects on colonic processes are of interest in colon can-
2-pentanone (Fig. 3). cer research.
In conclusion our results indicate that the flavouring com-
DISCUSSION pound 2-pentanone, found in food such as banana and
carrot,28,31) inhibits the COX-2 enzyme in vitro. It may there-
A daily intake of fruit and vegetables is believed to reduce fore influence the anti-inflammatory and chemopreventive ef-
the cancer risk, as it contains compounds with chemopreven- fects of fruit and vegetables. Further studies on flavouring
tive properties.15,24) Many dietary phytochemicals possess compounds as chemopreventive agents would be of value in
anti-inflammatory properties, whereof the COX-2 inhibitory the field of colon cancer research and may contribute to fu-
effect is one of the most studied. ture dietary recommendations.
March 2008 537

Acknowledgements Supported by grants from Agricul- 24) Howe G. R., Benito E., Castelleto R., Cornée J., Estève J., Gallagher
tural Sciences and Spatial Planning (FORMAS), Swedish R. P., Iscovich J. M., Deng-ao J., Kaaks R., Kune G. A., Kune S.,
L’Abbé K. A., Lee H. P., Lee M., Miller A. B., Peters R. K., Potter J.
Cancer Society, and Stiftelsen Ruth och Richard Julins Fond. D., Riboli E., Slattery M. L., Trichopoulos D., Tuyns A., Tzonou A.,
This work was partly supported by ECNIS (Environmental Whittemore A. S., Wu-Williams A. H., Shu Z., J. Natl. Cancer Inst.,
Cancer Risk, Nutrition and Individual Susceptibility), a net- 84, 1887—1896 (1992).
work of excellence operating within the European Union 6th 25) Ringbom T., Segura L., Noreen Y., Perera P., Bohlin L., J. Nat. Prod.,
Framework Program, Priority 5: “Food Quality and Safety” 61, 1212—1215 (1998).
26) Ringbom T., Huss U., Stenholm Å., Flock S., Skattebøl L., Perera P.,
(Contract No. 513943). Special thanks are due to Uddeholms Bohlin L., J. Nat. Prod., 64, 745—749 (2001).
Aktiebolags stipendiefond for the grant to Jenny Pettersson. 27) Karlsson P. C., Huss U., Jenner A., Halliwell B., Bohlin L., Rafter J. J.,
The authors also want to thank Dr. Ulrika Huss for helpful J. Nutr., 135, 2343—2349 (2005).
discussion. 28) Nogueira J. M. F., Fernandes P. J. P., Nascimento A. M. D., Phytochem.
Anal., 14, 87—90 (2003).
29) Zabetakis I., Holden M. A., J. Sci. Food Agric., 74, 421—434 (1997).
REFERENCES 30) Winter M., Willhalm B., Helv. Chim. Acta, 47, 1215—1227 (1964).
31) Linko R. R., Kallio H., Pyysalo T., Rainio K., Z. Lebensm. Unters.-
1) Miller S. B., Semin. Arthritis Rheum., 36, 37—49 (2006). Forsch., 166, 208—211 (1978).
2) Lee S. H., Soyoola E., Chanmugam P., Hart S., Sun W., Zhong H., 32) Mariaca R. G., Fernández-García E., Mohedano A. F., Nuñez M., Food
Liou S., Simmons D., Hwang D., J. Biol. Chem., 267, 25934—25938 Sci. Tech. Int., 7, 131—139 (2001).
(1992). 33) Pudil F., Víden I., Velísek J., Davídek J., Z. Lebensm. Unters.-Forsch.,
3) Smith W. L., Dewitt D. L., Adv. Immunol., 62, 167—215 (1996). 177, 181—185 (1983).
4) Weaver S. A., Russo M. P., Wright K. L., Kolios G., Jobin C., Robert- 34) Pinnel V., Vandegans J., J. High Resol. Chromatogr., 19, 263—266
son D. A., Ward S. G., Gastroenterology, 120, 1117—1127 (2001). (1996).
5) Dimberg J., Samuelsson A., Hugander A., Söderkvist P., Gut, 45, 35) Gold L. S., Slone T. H., Ames B. N., “Food Safety and Food Quality.
730—732 (1999). Issues in Environmental Science and Technology 15,” ed. by Hester R.
6) Subbaramaiah K., Dannenberg A. J., Trends Pharmacol. Sci., 24, 96— E., Harrison R. M., Royal Society of Chemistry, Cambridge, 2001, pp.
102 (2003). 95—128.
7) Thun M. J., Henley S. J., Patrono C., J. Natl. Cancer Inst., 94, 252— 36) Feng X. M., Ostenfeld Larsen T., Schnürer J., Int. J. Food Micribiol.,
266 (2002). 113, 133—141 (2007).
8) Sinicrope F. A., Molecular Carcinogenesis, 45, 447—454 (2006). 37) Ostenfeld Larsen T., Svendsen A., Smedsgaard J., Appl. Environ. Mi-
9) Chen W.-S., Liu J.-H., Wei S.-J., Liu J. M., Hong C.-Y., Yang W. K., crobiol., 67, 3630—3635 (2001).
Cancer Sci., 94, 253—258 (2003). 38) Lewis H. L., J. Gen. Microbiol., 63, 203—210 (1971).
10) Evans J. F., Am. J. Clin. Oncol., 26, S62—S65 (2003). 39) Fadda S., Lebert S., Leroy-Sétrin R., Talon R., FEMS Microbiol. Lett.,
11) Arber N., Eagle C. J., Spicak J., Rácz I., Dite P., Hajer J., Zavoral M., 210, 209—214 (2002).
Lechuga M. J., Gerletti P., Tang J., Rosenstein R. B., Macdonald K., 40) Statheropoulos M., Sianos E., Agapiou A., Georgiadou A., Pappa A.,
Bhadra P., Fowler R., Wittes J., Zauber A. G., Solomon S. D., Levin B., Tzamtzis N., Giotaki H., Papageorgiou C., Kolostoumbis D., J. Cro-
N. Engl. J. Med., 355, 885—895 (2006). matogr. B, 822, 112—117 (2005).
12) Smigel K., J. Natl. Cancer Inst., 92, 297—299 (2000). 41) Schiffman S. S., Bennett J. L., Raymer J. H., Agr. Forest Meterol., 108,
13) Dogne J. M., Hanson J., Supuran C., Pratico D., Curr. Pharm. Des., 12, 213—240 (2001).
971—975 (2006). 42) Schnell S., Wondrak C., Wahl G., Schink B., Biodegradation, 2, 33—
14) Heber D., J. Postgrad. Med., 50, 145—149 (2004). 41 (1991).
15) Lampe J. W., Am. J. Clin. Nutr., 78 (Suppl.), 579S—583S (2003). 43) Van Lier J. E., Da Costa A. L., Smith L. L., Chem. Phys. Lipids, 14,
16) Surh Y.-J., Nat. Rev. Cancer, 3, 768—780 (2003). 327—335 (1975).
17) Perera P., Ringbom T., Huss U., Vasänge M., Bohlin L., “Bioactive 44) Ohnishi S., Shibamoto T., J. Agric. Food Chem., 32, 987—992 (1984).
Compounds from Natural Sources,” ed. by Tringali C., Taylor & Fran- 45) Calviello G., Di Nicuolo F., Gragnoli S., Piccioni E., Serini S., Mag-
cis, London, 2001, pp. 433—472. giano N., Tringali G., Navarra P., Ranelletti F. O., Palozza P., Carcino-
18) Goel A., Boland C. R., Chauhan D. P., Cancer Lett., 172, 111—118 genesis, 25, 2303—2310 (2004).
(2001). 46) Yamada M., Niki H., Yamashita M., Mue S., Ohuchi K., J. Pharmacol.
19) Subbaramaiah K., Chung W. J., Michaluart P., Telang N., Tanabe T., Exp. Ther., 281, 1005—1012 (1997).
Inoue H., Jang M., Pezzuto J. M., Dannenberg A. J., J. Biol. Chem., 47) Klinder A., Gietl E., Hughes R., Jonkers N., Karlsson P., McGlyn H.,
273, 21875—21882 (1998). Pistoli S., Tuohy K., Rafter J., Rowland I. R., van Loo J., Pool-Zobel
20) Huss U., Ringbom T., Perera P., Bohlin L., Vasänge M., J. Nat. Prod., B. L., Int. J. Cancer Prev., 1, 19—32 (2004).
65, 1517—1521 (2002). 48) O’Brien J., Wilson I., Orton T., Pognan F., Eur. J. Biochem., 267,
21) Elango E. M., Asita H., Nidhi G., Seema P., Banerji A., Kuriakose M. 5421—5426 (2000).
A., J. Appl. Genet., 45, 469—471 (2004). 49) Bradford M. M., Anal. Biochem., 72, 248—254 (1976).
22) Kim S. S., Oh O.-J., Min H.-Y., Park E.-J., Kim Y., Park H. J., Han Y. 50) Laemmli U. K., Nature (London), 227, 680—685 (1970).
N., Lee S. K., Life Sci., 73, 337—348 (2003). 51) Nordling M. M., Nygren J., Bergman J., Sundberg K., Rafter J. J.,
23) Prashar A., Locke I. C., Evans C. S., Cell Prolif., 39, 241—248 Chem. Res. Toxicol., 15, 1274—1280 (2002).
(2006).