Photoactive furocoumarins in fruits of some umbellifers

Phykxkmurry. Vd 26. tie I. pp 165 169. 1987 Pnntedm Grat Bnwn

PHOTOACTIVE

0031-9422.87I3 00 + 0.00 PcrpmonJoumah Ltd

FUROCOUMARINS

0. C~KA,

IN FRUITS

S. K. CHAUDHARY. P. J. WARRINGTON and

Lkpanmcn~of Biology. Umvcrsityof Vrrona.

OF SOME UMBELLIFERS M. J. ASHWOO&MITH*

P.O. Bar. 1700, Victoria, B.C., V8W 2Y2 Canada

(Receiwd 2 Apil 1986) Key Wor4 l&x--Umkllifcrae; fruits; photoactIve furocoumarlns; HPLC arulyir photobtolo84 assay. quantification; psoralcn: S-mcthoxypsoralcn; 8-mcthoxypsoralcn; isopimpincllin: lsolmpcrstorm; rmperatorm.

oxypeuccdanrn: sphondin; chemotaxonomy. Ahtract-Fruits of Anefhwn graceolens, Angelica archangclica, Anthriscw cerefolium, Apiw grareolens. Carum carai, Corkandrum saticum, Cuminum cyminum, Daucus carota, F~niculrun vulgare, Herackwn sphondpliwn, Lerisricwn oficinale, Pasfinaca sariva, Petroselinw crispwn and Pimpinella aniswn were analyscd for photoactive furocoumarms using HPLC analysis and photobiological bioassay. Thcsc scnsitivc methods rcveakd furocoumarins not previously detected. _.--- __

INTRODL’ClIOS

RESLLTSAlvD DISCUSSION

The Umbclliferae is the family richest in species containing furocoumarins [ 1,2]. Some furocoumarins such as psorakn (l), S-rncthoxypsorakn (S-MOP, 2) and 8methoxypsorakn (8-MOP, 3) are potent photosensitizers when activated by near-UV light (3t%38Onm). They intercalatc readily into DNA and form light-induced mono- or di-adducts with pyrimidinc bases. Thus they are phototoxic, mutagenic and photocarcinogcnk. Severe dermatitis can result after contact with furocournarincontaining plants in tbc prcscna of sunlight [3]. Consumption of furocoumarins may also prcscnt some toxicological risk to man. In PUVA (psorakn-UVA light) photocbcmothcrapy for psoriasis about 20 mg of 8-MOP is administacd daily. This treatment is, however. recognizcd by the World Health Organization to be causally r&cd to human skin cancer 143. Various parts of many umbcllifcrous plants have been utilized by man for centuries [S]. la recent years thcrc has been dietary concern about the furocournarin content of umbcllifcrous vcgctabks such as akry [6], parsnip 173 and carrot [8,9]. It is known, however, that the highest conantration of these mokcuks is found in the fruits. Quite a few umbcllifcrous fruits, and the oils obtained from them, are widely used for flavouring in foods and beverages. Some are used in pcrfumcry. The fruits of hogwad, Herocleaun sphondylium L., are made into an alcoholic beverage in some European and Asian arcas and used for a liqueur in France [S]. The medicinal value of umbcllifcrous fruits, once highly prixcd, is kss important today. Only a few umbclliferous fruits have been recently analysed in detail and the photoactive furocoumarins quantified [IO. I I]. Using HPLC analysis [IZ] and an ultrasensitive bioassay [ 131, we report in this paper on tbc concentration of photoactive furocournarins in the umbclliferous fruits most frcqucntly usal by man.

The concentrations of photoactive furccoumarins 1 IO in various umbclliferous fruits are prcscntcd in Table 1. The plant specks are arranged according to their furocoumarin kvcl. The ultrascnsitivc bioassay was used for the detection of photoactive furocoumarins [ 133. The lower limits of the bioassay dctcction for compounds such as impcratorin (6) and sphondin (IO) were not previously known and have been cstablishcd They are impcratorin I x IO-’ g and sphondin 5 x low9 g. Also included in the tabk is isopimpincllin (4). which is considered nonphototoxic [3] but its phototoxicity has lnzn dcmonstratcd by Ivie [ 141 and Kavli et al. [IO]. The phototoxicity of isopimpincllin is weak, however, and cannot be detected by the bioassay even at a concentration of 4 x IO’ ‘g. The fruits of Angelica urchangelica L. had the highest content of total photoactive furocoumarins. approximately 1.29% dry weight. The total coumarin content in ripening fruits was reported to reach up to 2.59% then to decrease to 1.46 % in ripe fruits [IS]. All five furocoumarins detected in our sampk were reported previously 123. We failed to detect angclidn (9). also rcportcd in fruits of angelica [ 163. The fruits of parsnip, Pasrinoca sarrra L. had considerably lower amounts of photoactive furocoumarins (co 0.060.19% dry weight) than angelica. There is obviously great variation in furocoumarin content and composition in various parsnip cultivars. Orlov and Sire&o [IA found the total furocoumarin content in fruits of 20 varieties of cultivated parsnip to range from 0.12 to 2.63% dry weight. 8-MOP was the major component in eleven varieties, impcratorin in eight varictics and S-MOP in one variety. Ecrenbaum [ 1I] reported the avcragc furocoumarin content in ripe wild parsnip seed to be approximately I y. of the dry weight. almost three tima greater than that of the seed of the cultrvar. In her study, all seven cultivars and the majority of individual wild plants had impcratorin as a major component. Altbougb we detected the sanu furocoumarins as Bcrcnbaum [I 11, that was a great variation in concentra-

*Resent oddrc9: Bourn Hall Clinic, Bourn Hall. Bourn,, Cambr~@, CB3 7TR. U.K. I65

Table I. Photoactive furocoumarin Eoncentrauons m vanow umbdlifaous

Plant

.--_

.- _ .

1

l

2

3

Compound+

--

4

5

fruits (values are IO rg/g fruit) Unknown

__.-.-

.- phorowAve

6

-

7

10

compounds

-

Angeka

Hab. Parsnip Hollow Crownb All American’ Harris Mc&lC Hopw=d Wildd Parsky Champion Mou Curledb Singk Hardy Mianb Grccll Velvet’ Clivte Hamburg Roolcd’ Cekriac Gun! Smooth Prague’ Giant Smooth PragucJ Cekry Pans Gokknb Utah Tall Grant LoMF Hab’ FCUIUI Vcgetabk’ Herb’ ArUK Hab’

3477.00

-

429.76 402.94 213.48 136X

I.435 I.335 9.728

12.44p lo.sOg 9.22 Q ss9g I.§

128.20 3852.00

5030.00

.-

-

.-

_-

35576 758.60 161.16

.-

7.61 3.14 3.01

2 2 2

.-

:

4

642.10 205.16 682.14 108.90 169.98 91.44 25.~3 :. § :.§ 5.375 R§

I I.16 16.93

-

..

3.1gg

6.38

:

:

5.24 !

I.5 I.5

.-

I

I.§

I

b§

__

-

I

l.8

-

-

I.§

I.§

-

I.§ I.0 I.8 I.§ I.§ I.§

I.$ I.8 I.8 I.4 I.§ I.0

-

-

-

-

5.62 0 5.005 3.520 :*o -

35.76 1.3gp 5.92 § 4.09 0 0.69 p -

-

25.71 5 21.41 6 6.644 10.82§ -

--. .-

-

: :

2.30 6.68

-

-

8.00

-

COkMkt

Habc Dill Habb Caraway Hab’ Carrot Danvm Half Long‘ Mokum Fl Hybrid’ Gold Pak‘ Kundulu# Namesa Chanlcnay~ Cumin Herb’ ChCWll Hab’

427.30

-

: 12.820

._.

-

2.805 :. §

$.P :.§

3 I

-

_-

-

-.

-

.-

-

lSupcrszripts indimte souraz a. Suffolk Habr. Organic Farmas & Growers Ltd, SuITolk,U.K.; b. A.E. McKenzie Co., Llh Brandon.

Maniroh w c. Pacific Northwest Seed Co. lnc, Vanon, B.C., Can&% d, Oxford, U.K.; c. Thompson B Morgan Ltd. Ipwrh. U.K.; f, Buckalkld’s Ltd, VPacouva. B.C, Cana& g, Island s&d Co. Ltd. Victoria, B.C., Canad% h OnUno seed CO.Ltd.. Waterloo, Oorpno. Canado; I, Suttons Sadr Ltd, Torquay. U.K. +Scc formulae. $ Approximately 0.5 wig (HPLC detection hmit - 0. I rug). QNcw mformation. lApproximately 0.005 rgig (biouspy detection lim11- 0.001 w/g for 5-MOP and 8-MOP).

tion of the individual compounds in our samples. Two weakly active unknown photoactive compounds were also present. The fruits of hogwaxi, Heradewn spkodylium L., posses4 S-MOP as a major component. All identified

furocoumarins san in our sampk have been reported previously [2]. We were unable to identify one photoactive compound which was present in moderate concentralion. There were also trace amounts of three other photoactive compounds. Angelicin and psorakn, rep-

Phot~tivc

7

Oxypcuc~dmin

furoa3umarins in fruits of some umbcllifers

A”\

0CH7Ctl--Ma~

167

H

orted by Karlscn et al. [ 183and by Mandcnova er al. [ 191, could not be detcctcd in our sample. The fruits of parsky. Petrowlinum crispum L. cultivars grown for kavcs, had a total furocoumarin content of approximately 0.0037 of the dry weight. The major component was oxypeucedanin; other furocoumarins detected wcrc S-MOP, 8-MOP. isoimpcratorin (5) and impcratorin (6) The fruits also contained a small amount of paorakn, normally a relatively rare component in umbclliferous fruits, According to Murray cr a/. [2], only some specks of Hera&urn, Foeniculum, Pran~os and Trachyspermum contain psorakn in tbc fruit. The fruits of the cultivar Hamburg Rootcd had an exceptionally low furocoumarin content. Kato et al. [ZO] reported a compound they idcntiticd as hcraclenol from parsky fruits From their NMR data it is obvious that this compound was in fact oxypcuadanin hydrate (8) which originated from oxypeucedanin (by the opening of the cpoxide ring [Z]) during the hydrolysis in 2 N hydrochloric acid. The fruits of akry and akriac, Apium gratwolens L var. duke (Miller) Pers. and A. grweolens L.. var. tapaceum (Milkr) Gaud.. respectively, showed a similar pattern, 5-MOP being slightly higher in akriac. Garg er a/. [21] isolated isoimpcratorin as well as 5-MOP and isopimpincllin from a large sampk (4 kg) of akry fruits The fruit of lovage, ~isricum oficinale Koch, contained impcratorin as a major component and small amounts of S-MOP and psorakn. The bioassay also confirmed tbc prcscnce of 8-MOP. Nava [22] and Dauksha and Dcnisova 1233 isolated 5-MOP from Lovaga fruits. Psorakn was character&d together with S-MOP by Karlscn et d. [24] as being present in the root. The fruits of two varieties of fatal, Fcxniculum a&are Milkr, were analyscd. The fruits of vegetabk fcnncl, F. wlgure Milkr subsp. duke (IX.) RcrtoL. contained small amounts of S-MOP and impcratorin and trrcts ( < 0.5 &g) of psorakn and isopimpincllin. There was

also one unknown, weakly photoactive compound. The fruits of the herb. F. vulgare Miller subsp. vu/gore, had the same furocoumarins but only in tract amounts. 5-MOP and psorakn have been reported in the fruits of fennel by Menda and Pocciro (in Murray er a/. [3], unpublished results). The fruits of anise, Pimphello anisum L, coriander, Coriundrum sotiuum L, dill, Anerhum grabeolens L., caraway, Canun carui L., and carrot, Daucus carota L. had no detectable furocoumarins by HPLC analysis. Howcvcr. using the ultrasensitive bioassay, we were able to detect traces of S-MOP and 8-MOP in all of them. Roth compounds were at the limits of the detectton (1 x 10m9g [ 133) and their kvels in the fruits could bc atimatcd only approximately as c 0.005 ccg/g. The fruits of anise also had two other unknown, slightly photoactive compounds. S-MOP has been rcportcd from anise fruits by Kartnig and Scholz 1253. The fruits of coriander had two additional weakly photoactive compounds, one unknown and the other probably impcratorin. Kartnig [26] rcported the presence of S-MOP in coriander fruits. S-MOP was also found in dill fruits by Kartnig [26] and Dranik and Prokopcnko [27]. There is oo reference to the prcscncc of the furocoumarins in the caraway fruits or in any other parts of the plant. Small amounts of S-MOP and 8-MOP were detectad rcccntly in kaf surface wax [28] and in leaf and root of the carrot [9]. The fruits of cumin, Ctinw cyminum L., and chervil, Anhrixus cere/oIium (L.) Hoffm., did not contain any photoacdn furocoumarins, as shown by both HPLC analysis ami bioassay. There are no reports of photoactive furocoumarins in any part of these plants. The total concentration of photoactive furocoumarins is extremely high in the fruits of angclica and parsnip. It reaches over 12 mg/g in angelica and over I me/g on average in parsnip. The furocoumarin content in other analyscd umbcllifcrous fruits is much Iowa or negligibk.

0. CMA cr al.

168

In nature many umbcllifcrous fruits containing seeds remain in the soil for several months before germination; the prcscrkc of furocoumarins could be related to their possibk rok as antibacterial, antifungal, alklopathic agents or seal dormancy regulators 1291. The most common compounds found in umbclliferous fruits are SMOP and 8-MOP. both of which are strongly photoactive. Sad furocoumarins were usal by Crowdcn er 01.[ 301 as taxonomic markers in their chcmotaxooomic survey of Umbclliferae. From a chcmosystematic point of view, the absence of a certain furocoumarin is as important as a positive detection. However, many compounds can occur in the plant in amounts which are below the detection limits of the analytical methods. The introduction of a new. more sensitive analytical method for the detection of furocoumarins cnabkd us to rc-cvaluate earlier obscrva(ions. New information on the presence of various furocoumarins in umbclliferous fruits are presented in Table I. The presence of 8-MOP and S-MOP in carrot leaf wax, kavcs and root [9,28] was also confirmed in carrot fruits. From a taxonomic point of view, this new finding suggests the close relationship of Doucus with other members of subfamily Apioideac. where furocoumarins are common. This is also true for the caraway which belongs to the tribe Apiece but no previous record on the presence of furocournarins in caraway fruits exists. The fruits of only two specks had no detectable furocoumarins by our analysis. Chervil belongs to the tribe Scandiceae and cumin IO the tribe Apkac; furocoumarins arecommon in both tribes. II is possible that the sample analyscd was too small for detection.

Standards

Fwommarm CO..

srmdards.

was

U.S.A.

Pharmazutika

syo~hcsixod in our laboratoncs. Roth

Chcrmcal

lsopimpinclhn Chemistry,

Co..

Carl

purchased

&MOP

from Sigma Chemical Co.. St. Louis. MO, gift from Gcrot

Time Fig.

1. Sepuatwn

phase HPLC.

lsoimpcratorin

&tails

of HPLC

MOP.

4 - rsoptmpincRi~

conditions.

1 ml/mm

dissolval in CHCl,.

was obtained from

200@ml.

See Fig. 1 for

Karlsruhc.

analysu, Merck praoatcd

KG,

F.R.G.

Columbir,

of

Vancouver. duwronii.

sphondin from the roots of Herculcum lana~um,and oxypeucd-

P&IU nuunal.

identityof

of Prrroselinum crispum [ 3 I]. The

the tasasl three compounds Fruits

was atablishal

by NMR

and MS.

U.K. Depending on the furccounurin

were colkctcd in Oxford, cootmt cxpeacd

of 50. IO0 and 500 mg of fruits were extrmztal twia EtOAc for 24 hr. Thu extraction no furocoumarins HPLC

extracts

and broassay. The combined

sampks

as dctcrmmcd

by

extracts were cvaporaIed.

dissolved m 0.5 ml CHCl,,

filtered through a 0.5 pm filter and

dtrectly analyscd by HPLC

and TLC.

HPLC and TLC. All solvents were HPLC from Burdick and Jackson Laboratories. HPLC

Si-5

(4 mm x 3Ocm)

3390 A integrator.

was used for

HPLC

A Varian analysis

employing the method of ref. [ I I] w?tb a slight mod&cation. solvent wascydohcxa.noico-Pr,O a Bow rate of I ml/nun. separated with l5:b

The

n+rnyl alcohol (IS:4:05)with

cxoepr for uoimpcratorin

EtOAc

was

io cyclobexar~

at

3 - a-

254 nm

and

were

&nbpcd.

the

Kpentbn

of standards

silica gd K 60 &IS

first

hcuotpcn~EtOAc with UV (3C&380

AT = 0.16auf.

were used at concns of 5. JO and

CXXnt indicator) were used lad 1wodimcnsioMl with

(7:7:6).

CHCl,

TLC fluor-

chromatograms

and

Vnualizatron

For

(without second

with

was performed

nm) light.

Phorobio/qka/

puoy.

repair d&ient

- ),which

This

ultrascnsnin

alkyhtiag

mutant, Erchrichia

is exrmncly

photobiobgical

coli Bs-J (ret +, exr - ,

hrc

sensitive to UV radiation and to chemical

agents [33].

was used as a test micro-organism.

Acknowkdt7emenrs-Thu M. J. Ashwood-Smith Canada

and the

which was

with a flow rate of

rcsoarch was supported from the Medial

Natural

by grants to

Research CourxciJ of

scicncos pnd Engineering

Rcaauch

Council of Canada.

grade, purchased

system equipped with a Rbcodyne 7125 loop in&ctor was

Column

Detoctioo

REFERENCES

U.S.A. A Vanan 5000

used with detection by UV at 254 run. Data proccssiog was done by peak area on a Hewktt-Packard

2 = LMOP.

7 - oxypeuczdanin,

with 50 ml

was sulficient smcz there were

in subqumt

normal

text for full

assay has bc+n described in detail rn rcfs. [9, 13.321. The DNA

were purchxscd locally and from the

U.K. Fruits of Herac&um splmdyfiwn

by

Sa

10 - rphondm.

Standa&

Bcitish

1 - Puuakn.

6 - imvtorin,

was a

lmpcratonn

atanduds

from

U.S.A. S-MOP

was isolated from the roots of An&co

anm from the laws

furocoumum

The injoztioo volume was 20&

Vienna. Austria. Angelicin was Roth

of

of

(Mln.1

was purchMcd

was a gift from Dr. J. P. Kutncy. Dcpprtmen~ University

I

9 - an&in,

Psorakn

MI,

Kakxmaxoo.

ug/ml

4 9

EXPCRIMEhTAL

UpJOhn

40

Nicbcn. B. E (1970) Coumar uu o/ umbellijtrcur Phus. nc Royal hoisb School of Pharmacy. Copcnhapo. Murray, R. D. H, Mender. J. and Brown, S. A. (1982) 7%~ Narural

Coummfnr.

John Wiky.

The Stone Bridge Press,

BrislOl. Pathak.

M. A. (1974) io Swlighr

Ha&r.

L C.. Sciji. M. and Kukita.

consulung at.)

p. 495.

Univmicy

and Man (Pathak, A.. cds.; Fitzpatrik. of Tokyo

M. A.. T. B..

Press, Tokyo.

Photoactlvc

4. IARC

(1983) Chemicals. lndusrriol

Asrcwuucd

wirh

Evaluatmn

of

Cancer the

Process

in Hnmnt.

Carcinogenic

furcxouma rins In frulls of some umbcllifcrs

ad

Industries

Monograph Risk

of

Humans, Supplement 4. Vol. I. p. 29. IARC.

Cauwet-Marc,

on the

Chcmkak

Mob.

Lyon, France.

Mw

6. Bclcr. R. C.. IYK. G. W, Oertli. Food Chcm. Taxi.

E. H. and Holc. D. L. (1983)

20. Kate,

7. Ivie. G. W.. Hoh, D. L. end Ivcy. M. C. (1981) S~ICKC 213.

21. Garg,

909. W.. B&r.

R. C. and Holr. D. L. (1982)1.

AgrK.

Food

Chem. 30.413. 9. Ceska, 0.. Chaudhary.

S. K., Warrington

Smith, M. J. (1986) PhyrcKhermsrr) IO. Kavli. G.. K&an.

H, Midclfan.

(1983) Photobiochcm Bcrcnbaum.

12. Chaudhary.

2S. 81.

Phofobiophys.

13. AshwoodSmlth.

S. K..

0..

IS. Zoclkov.

Yu

Khun.-Fcllm.

the

Musouri

Sasaki.

N,

S. R. and

Kltanava,

Y. and

17. 158. Sharma,

N.

D.

E. K. (IW9)

BIO/. XOI&I

L. E. J. and Baerhaim-Svcndsen,

38, I 13.

26. Kanmg.

T. (1966) ~rrrr.

27. Dranik.

L. 1. and Prokopcnko.

.&din.

S&n.

Ansrrwhm. 68. 131. A. P. (1969)

Khim.

28. Stadkr.

E. and Buser. H. R. (1985) Exprricnrla

40. 1984.

30. Crowden.

R. K.. Harbornc,

Phyrochemurry

Grs. 81.

L. Y. (1969) Rauit.

Rcsw.

No. 5.

32

Ashwood-Smith Warrmgton.

445. J.. Algcra.

M. A W

and

Bacrhaim-Svcndscn.

A.

I. P.. Carbonnier.

J., Grbonnier-Janau.

M. C..

M.

V. H. (1969)

B. K. (1969) Can. 1. Chcw~. 47. 2425 1.. Cc&a.

P. J. and Woodcock.

0..

Chaudhary,

S. K..

P. (1986) J. Chtm. Ecol

IZ

915. 33. lgali. S., Bridga,

See. Phurm. 37.249.

J. B. and Heywd.

Hbr.

a 1963.

31. Srcck. W. and Baiky. Y. E and Sircnko.

Prir.

5.437.

Inst. 26. I41 Phum.

A.

T. and Scholz, G. (1969) Ferrr. Seven, Ansrricbn. 71,

M. and Ro?m~svc~g. P. E. (1968) Tr. L*ningr. DtKh.

(1979)

Ia 1580.

( I968) Mcdd.Nor. Fmm.SCM.30. 169.

Liu. M.

289.

19. Mandenovq

from

29. Carbonnier. 1. and Molho, D. (1978) Bull. Mus. Nar/ Nat., Ser. Phys.Chim. 19. 17.

16. Corcillius, F. (19%) Arch. Phurm. &r.

18 Karlscn.

Botany

Sysrmafqw.

276.

P. J. and Ashwood-

Phorobiol.

M..

Gupb,

Ombelli_ftres.

Irr

p. 675.

N. (1978) Phyrochcmisrry

25. Kanmg.

14. Ivie. G. W. (1978) 1. Agric. Feud Chem. 26. 1394.

(lW9)

Kobayashi.

a

22. Navcq Y. R. (1943) He/u. Chum. Acra 5. 1281.

23. 1809.

M. J., Poulton, G. A., Cc&a.

and Fumlss. E_ (1983) Phomhtm.

VoL6.

24. Knrlscn. J, Boomsma.

5. 159.

Warrington.

Systematic

23. Dauksha. A. D. and Denisova, (Alma-Ara) No. 4. 88.

K.. Voldcn. G. and RM. J.

Xl. R. (1984) Phyrdwmurry S. K.,Caka.O..

P. J. and Ashwood-

Smith. M. 1. (1985) J. Agric. Food C’hem. 33. 1153.

17. Orlov.

f,

Phymdwmis~y

8. Ivic.G

1I.

Guden,

Takabhi,

21, 163.

1. P. (1982) in hs

piurbdisciplinaires in

Boclaial

5. French, D. H. (1971) Bar. 1. Unn. Sue. 64 (Suppl. I). 385

M. T., Guyot, M..

A. M.. Ccraau-brrinl,

D. and Rcduron.

COltrribUriCflS

to

169

8. A.. A.&wood-Smith,

(1970) .Hurar. Rex 9. 21.

M. J. and Scott, B. R.