Biol. Pharm. Bull. 26(9) 1371ﾑ1373 (2003)

September 2003 Notes Biol. Pharm. Bull. 26(9) 1371—1373 (2003) 1371

Inhibitory Effects of Citrus Fruits on Cytochrome P450 3A (CYP3A) Activity in Humans

,a a a a a Ken-ichi FUJITA,* Muneaki HIDAKA, Norito TAKAMURA, Keishi YAMASAKI, Tomomi IWAKIRI, a a b b Manabu OKUMURA, Hirofumi KODAMA, Masatoshi YAMAGUCHI, Tsuyomu IKENOUE, and a Kazuhiko ARIMORI a Department of Pharmacy, Miyazaki Medical College Hospital; and b Department of Obstetrics and Gynecology, Miyazaki Medical College Hospital; 5200 Kihara, Kiyotake-cho, Miyazaki-gun 889–1692, Japan. Received June 11, 2003; accepted July 3, 2003; published online July 4, 2003

The capacities of citrus fruits to inhibit midazolam 1Ј-hydroxylase activity of cytochrome P450 3A (CYP3A) expressed in human liver microsomes were evaluated. Eight citrus fruits such as ama-natsu, banpeiyu, Dekopon, hassaku, hyuga-natsu, completely matured kinkan (Tamatama), takaoka-buntan and unshu-mikan were tested. We also examined the inhibition of CYP3A activity by grapefruit (white) and grapefruit juice (white, Tropicana- Kirin). The addition of a fruit juice prepared from banpeiyu, hassaku, takaoka-buntan or Tamatama caused the inhibition of the microsomal CYP3A activity. The inhibition depended on the amount of a fruit juice added to the incubation mixture (2.5 and 5.0%, v/v). The fruit juice from banpeiyu showed the most potent inhibition of CYP3A. The addition of a banpeiyu juice (5.0%, v/v) resulted in the inhibition of midazolam 1Ј-hydroxylase activity to about 20% of control without a fruit juice. The elongation of the preincubation period of a fruit juice from banpeiyu (5.0%, v/v) with the microsomal fraction (5 to 15 min) led to the enhancement of the CYP3A inhibition (5% of control). Thus, we discovered ingredients of banpeiyu to be inhibitor(s) or mechanism-based inhibitor(s) of human CYP3A activity, but the inhibitory effects of them were somewhat lower than those of grapefruit. Key words banpeiyu; Dekopon; hassaku; hyuga-natsu; kinkan (Tamatama); takaoka-buntan

Grapefruit juice has been taken by many people living in So far, there has been no report on the effects of these cit- all over the world. In early 1990’s, grapefruit juice was docu- rus fruits on human cytochrome P450 3A (CYP3A) activity. mented to affect the oral bioavailability of felodipine.1,2) The Therefore, in the present study, we investigated whether mean bioavailability of felodipine administered with grape- the components present in the citrus fruits inhibited the fruit juice was more than 250% that with water. The diastolic CYP3A activity. The ability of ingredients of the fruits to blood pressure was lower and the heart rate was higher inhibit the midazolam 1Ј-hydroxylase activity of human than those seen with water. Adverse experiences, mainly CYP3A was examined by using human liver microsomes. accounted for by headaches, facial flushing and lightheaded- ness, were reported to be more frequent after intake of grape- MATERIALS AND METHODS fruit juice than water. Subsequently, various drugs which were orally administered have been proven to interact with Chemicals Glucose 6-phosphate, glucose 6-phosphate grapefruit juice.3—8) These drugs differ in their chemical dehydrogenase and NADPϩ were obtained from Oriental structure and pharmacological properties, but are commonly Yeast (Tokyo, Japan). 1Ј-Hydroxymidazolam and midazolam metabolized by a form of cytochrome P450,9) namely were from Wako Pure Chemicals (Tokyo, Japan). Pooled cytochrome P450 3A4 (CYP3A4). Studies performed by human liver microsomes were purchased from Daiichi Pure Fukuda et al.,10,11) Guo et al.12) and Tassaneeyakul et al.13) Chemicals (Tokyo, Japan). All other chemicals and solvents have shown that the furanocoumarin derivatives identified were of the highest grade commercially available. from grapefruit juice strongly inhibited the catalytic activity Samples of Fruits and Juice Citrus fruits, ama-natsu, of CYP3A4, and caused the decrease of the first pass metab- banpeiyu, Dekopon, hassaku, hyuga-natsu, completely ma- olism of orally administered drugs by CYP3A4. Subsequent tured kinkan (Tamatama), takaoka-buntan, unshu-mikan, studies have demonstrated that grapefruit juice resulted in the grapefruit (white), grapefruit juice (white, Tropicana-Kirin) loss of intestinal CYP3A4 protein without changing the spe- and navel orange were obtained from local commercial cific mRNA levels.14,15) On the other hand, common orange sources and their available information is shown in Table 1. juice has been documented that it is incapable of inhibiting All the fruit samples and grapefruit juice sample were tested the catalytic activity of CYP3A4.2) Taking these results into soon after they were squeezed and filtered or the juice pack- account, the inhibitory effects of citrus fruits appeared to de- age was opened. pend on the fruit species because of the difference of the Assay of Midazolam 1Ј-Hydroxylase Activity of Human 2,12) components contained in each of a citrus fruit. CYP3A A typical incubation mixture consisted of 100 mM Miyazaki prefecture is located at the southern part of sodium potassium phosphate buffer (pH 7.4), 50 m M EDTA ϩ Kyushu Island in Japan. A vast array of citrus fruits have disodium salt, an NADPH-generating system (0.5 mM NADP , been produced in Miyazaki area because of its warm climate. 5mM MgCl2, 5 mM glucose 6-phosphate and 1 unit/ml glu- The citrus fruits are taken by many people living not only in cose 6-phosphate dehydrogenase), and microsomal fraction the area but also in the other regions around there, especially of the human liver in a final volume of 0.5 ml. The concen- in winter and spring seasons. tration of midazolam was 10 m M. The protein content and re- ∗ To whom correspondence should be addressed. e-mail: [email protected] © 2003 Pharmaceutical Society of Japan 1372 Vol. 26, No. 9

Table 1. Available Information about the Citrus Fruits Used in This Study

Citrus fruits Origin Fruit or juice Processing Japanese English Species

Ama-natsu Natsudaidai Citrus natsudaidai HAYATA Kumamoto Japan Fruit Squeezed Banpeiyu Banpeiyu Citrus grandis OSBECK Kumamoto Japan Fruit Squeezed Dekopona) Dekopon Shiranui mandarin Suppl. J. Miyazaki Japan Fruit Squeezed Hassaku Hassaku orange Citrus hassaku HAT. ex TANAKA Miyazaki Japan Fruit Squeezed Hyuga-natsu New summer orange Citrus Tamurana Miyazaki Japan Fruit Squeezed form orange Hyuganatsu Kinkan (Tamatamaa)) Kumquat Fortunella hindsii Miyazaki Japan Fruit Squeezed Takaoka-buntan Pomelo Citrus grandis Miyazaki Japan Fruit Squeezed Unshu-mikan Satsuma mandarin Citrus unshu Marcovitch forma Miyazaki Japan Fruit Squeezed Miyanaga-wase Grapefruit (white) Grapefruit (white) Citrus paradisi Florida U.S.A. Fruit Squeezed California U.S.A. Juice Reconstituted juice (Tropicana-Kirin) Navel orange Navel orange Citrus sinesis (LINN.) OSBECK var. California U.S.A. Fruit Squeezed brasiliensis TANAKA

a) A registered trademark. action time were predetermined based on linearity between RESULTS AND DISCUSSION microsomal protein concentration (up to 0.2 mg/tube) and the reaction time (up to 2 min) versus metabolite formation rate. Inhibition of midazolam 1Ј-hydroxylase activity of human Reactions were initiated by the addition of midazolam. After CYP3A by ingredients of various citrus fruits was examined. incubation at 37 °C for 2 min, 5 ml of ethyl acetate was added The inhibition of human CYP3A by a juice prepared from to stop the reaction. Five hundred picomoles of clonazepam each of a citrus fruit is summarized in Table 2. The addition was added as an internal standard. The mixture was extracted of a fruit juice prepared from banpeiyu, hassaku, takaoka- with ethyl acetate and centrifuged at 3000 rpm for 10 min. buntan or Tamatama caused the inhibition of the microsomal The organic layer was transferred to another tube and the CYP3A activity. The inhibition depended on the amount of a solvent was evaporated. The residue was dissolved in 10 mM fruits juice added to the reaction mixture. Among the citrus sodium acetate, methanol and acetonitrile (9:1:1, v/v) fruits, banpeiyu juice showed the most potent inhibition. The (200 ml). Analysis of the 1Ј-hydroxymidazolam metabolite addition of a banpeiyu juice (5.0%, v/v) resulted in the inhi- was performed by HPLC using a computerized HPLC sys- bition of midazolam 1Ј-hydroxylase activity of human tem (Shimadzu model LC10A series, Shimadzu, Kyoto, CYP3A to about 20% of control without a fruit juice. The in- Japan) equipped with a TSK-gel ODS-120T analytical col- hibition was weaker than that of grapefruit juice. The same umn (4.6ϫ150 mm; 4 mm; TOSOH, Tokyo, Japan). The mo- amount of grapefruit juice (Tropicana-Kirin) caused the 99% bile phase consisted of 10 mM sodium acetate, methanol and inhibition of human CYP3A activity. No clear inhibition was acetonitrile (9:1:1, v/v) for solvent A and methanol and observed with ama-natsu, Dekopon, hyuga-natsu and unshu- acetonitrile (2 : 1, v/v) for solvent B. The metabolite was sep- mikan. arated using a linear gradient of 70% to 50% solvent A, 0 to We examined whether the component(s) of citrus fruits in- 25 min, at a flow rate of 1.0 ml/min. Quantification of the hibited human CYP3A in a mechanism-based manner. If in- metabolite was performed by comparing the HPLC peak area gredient(s) was a mechanism-based inhibitor(s) of human monitored at 240 nm to that of the internal standard. CYP3A, we would expect the inhibition to be enhanced by Inhibitory Effects of Citrus Fruits on CYP3A Activity preincubation of the ingredient(s) with the microsomal The inhibitory effects of citrus fruits on CYP3A activity preparation. Results are shown in Table 2. The inhibition were investigated according to the method of Guo et al.12) potency of a fruit juice from banpeiyu or hassaku to the with minor modifications. Briefly, a citrus fruit juice (12.5 or CYP3A activity observed with 15 min preincubation was 25.0 ml) was dried with a concentrator. The reaction mixture, higher than that observed with the 5 min preincubation. The before the addition of midazolam, described above was residual activity seen with a fruit juice prepared from ban- added and the residue of the citrus fruits was resuspended peiyu (5.0%, v/v) was 5% that seen with control without a with a vortex mixer. After a preincubation of the mixture at fruit juice. The potency was somewhat lower than that ob- 37 °C for 5 min, the substrate midazolam was added. The re- served with grapefruit juice (1.7%). action was performed as mentioned above. The inhibitory ef- Judging from these results, we conclude that ingredient(s) fect of citrus fruit on midazolam 1Ј-hydroxylation was ex- of banpeiyu is inhibitor(s) or mechanism-based inhibitor(s) pressed as a percentage of the residual activity compared of human CYP3A activity, but the inhibitory effects are with the control. somewhat lower than those of grapefruit. Effects of Preincubation of Citrus Fruits on Human It is of interest to know what the component(s) of ban- CYP3A Activity As an index of a mechanism-based inhi- peiyu showing the inhibition or mechanism-based inhibition bition, a citrus fruit juice was preincubated at 37 °C for of human CYP3A is. It has been reported that the compo- 15 min in the reaction mixture according to the method men- nents of grapefruit which inhibited human CYP3A were ex- tioned above. tracted by ethyl acetate.12) We preliminarily examined September 2003 1373

Table 2. Effects of Components Contained in Citrus Fruits on Midazolam 1Ј-Hydroxylase Activity of Human Liver Microsomes

Residual activity (%)

Citrus fruits 12.5 mla) 25 mla)

5 minb) 15 minb) 5 minb) 15 minb)

Ama-natsu 90 90 69 55 Banpeiyu 24Ϯ3c) 11Ϯ2c) 20Ϯ6c) 5.0Ϯ1c) Dekopon 150 130 130 90 Hassaku 46Ϯ18c) 33Ϯ25c) 33Ϯ14c) 15Ϯ11c) Hyuga-natsu 79d) 50d) 59d) 38d) Kinkan (Tamatama) 42 31 35 25 Takaoka-buntan 42Ϯ1c) 43Ϯ10c) 30Ϯ10c) 22Ϯ5c) Unshu-mikan 160 110 140 120 Grapefruit (white)e) 12 7.2 2.0 0.7 Grapefruit juice (white)e) 2.0 2.8 1.0 1.7 Navel orangee) 91 83 64 68

Each examination was performed in duplicate. The control activity of midazolam 1Ј-hydroxylation by human liver microsomes determined in the absence of a citrus fruit juice was 1.92 nmol/min/mg prot. a) Amount of juice used in assays. b) Preincubation period. c) Values are presented as meanϮS.D. of triplicate examinations. d) Values represent the mean of duplicate examinations. e) Grapefruit, grapefruit juice and navel orange were positive and negative controls, respectively. whether or not the ingredients of banpeiyu inhibiting the Clin. Invest. Med., 12, 357—362 (1989). CYP3A activity were extracted by ethyl acetate. The ban- 2) Bailey D. G., Spence J. D., Munoz C., Arnold J. M. O., Lancet, 337, 268—269 (1991). peiyu juice (1 ml) was extracted by ethyl acetate (3 ml). The 3) Bailey D. G., Arnold J. M., Munoz C., Spence J. D., Clin. Pharmacol. equivalent amount of the extract (5.0%, v/v) was incubated Ther., 53, 637—642 (1993). with human liver microsomes. The data obtained by us indi- 4) Schubert W., Cullberg G., Edgar B., Hedner T., Maturitas, 20, 155— cated that the components of banpeiyu showing the inhibi- 163 (1994). tion of CYP3A were not necessarily extracted by ethyl ac- 5) Dcharme M. P., Warbasse L. H., Edward D. J., Clin. Pharmacol. Ther., 57, 485—491 (1995). etate. The residual activities of CYP3A were about 30% 6) Kupferschmidt H. H., Ha H. R., Ziegler W. H., Meiner P. J., Krahen- (5 min preincubation) and 24% (15 min preincubation), re- buhl S., Clin. Phamacol. Ther., 58, 20—28 (1995). spectively. It seems likely that the component(s) of banpeiyu 7) Hukkinen S. K., Verhe A., Olkkola T. K., Neuvonen P. J., Clin. Phar- having the inhibitory effects on human CYP3A was different macol. Ther., 58, 127—131 (1995). from that of grapefruit. 8) Benton R. E., Honig P. K., Zamani K., Cantilena L. R., Woosley R. L., Clin. Phamacol. Ther., 59, 383—388 (1996). Interestingly, when the fruit juice prepared from Dekopon 9) Nelson D. R., Koymans L., Kamataki T., Stegeman J. J., Feyereisen R., or unshu-mikan was added to the incubation mixture (2.5%, Waxman D. J., Waterman M. R., Gotoh O., Coon M. J., Estabrook R. v/v), the CYP3A activity increased to 150% or 160%. W., Gunsalus I. C., Nebert D. W., Pharmacogenetics, 6, 1—42 (1996). CYP3A4 has been reported to exhibit not only the ho- 10) Fukuda K., Ohta T., Oshima Y., Ohashi N., Yoshikawa M., Pharmaco- motropic cooperativity but also the heterotropic cooperativ- genetics, 7, 391—396 (1997). 11) Fukuda K., Ohta T., Yamazoe Y., Biol. Pharm. Bull., 20, 560—564 ity, which could influence drug metabolism and excretion or (1997). 16—20) bioactivation. Effectors such as a-naphthoflavone 12) Guo L.-Q., Fukuda K., Ohta T., Yamazoe Y., Drug Metab. Dispos., 28, caused the stimulation of CYP3A4 activity toward some sub- 766—771 (2000). strates.16) The heterotropic cooperativity may be clinically 13) Tassaneeyakul W., Guo L.-Q., Fukuda K., Ohta T., Yamazoe Y., Arch. Biochem. Biophys., 378, 356—363 (2000). significant due to the role, since it can play in enhancing 14) Schmiedlin-Ren P., Edwards D. J., Fitzsimmons M. E., He K., Lown K. drug–drug interaction. One of the emerging hypotheses of S., Woster P. M., Rahman A., Thummel K. E., Fisher J. M., Hollem- the mechanism of CYP3A4 cooperativity involves multiple berg P. F., Watkins P. B., Drug Metab. Dispos., 25, 1228—1233 substrate binding sites.21,22) A cooperative single-enzyme (1997). model with two binding sites where product can be formed 15) Lown K. S., Bailey D. G., Fontana R. J., Janardan S. K., Adair C. H., Fortlage L. A., Brown M. B., Guo W., Watkins P. B., J. Clin. Invest., either from the single-substrate-bound form or from the two- 99, 2545—2553 (1997). 21,22) substrate-bound form of the enzyme was described. The 16) Schwab G. E., Raucy J. L., Johnson E. F., Mol. Pharmacol., 33, 493— component(s) such as flavonoid(s) contained in Dekopon and 499 (1988). unshu-mikan may also simultaneously bind to the active site 17) Shou M., Grogan J., Mancewicz J. A., Krauze K. W., Gonzalez F. J., of CYP3A enzyme with midazolam and may responsible for Gelboin H. V., Korzekwa K. R., Biochemistry, 33, 6450—6455 (1994). 18) Shou M., Dai R., Cui D., Korzwkwa K. R., Baillie T. A., Rushmore T. the heterotropic cooperativity which enhances the enzyme H., J. Biol. Chem., 276, 2256—2262 (2001). activity. 19) Harlow G. R., Halpert J. R., J. Biol. Chem., 272, 5396—5402 (1997). 20) Harlow G. R., Halpert J. R., Proc. Natl. Acad. Sci. U.S.A., 95, 6636— Acknowledgment We thank Mrs. Masako Fujita to pre- 6641 (1997). pare citrus fruit juices. 21) Korzekwa K. R., Krishnamachary N., Shou M., Ogai A., Parise R. A., Rettie A. E., Gonzalez F. J., Tracy T. S., Biochemistry, 37, 4137—4147 (1998). REFERENCES 22) Shou M., Mei Q., Ettore M. W., Jr., Dai R., Baillie T. A., Rushmore T. H., Biochem. J., 340, 845—853 (1999). 1) Bailey D. G., Spence J. D., Edgar B., Bayliff C. D., Arnold J. M. O.,