New halometabolites from Caldariomyces fumago

Phytochemstry, Vol 27, No 4, pp 1093 1096, 1988 Prmted m Great Bntam

0031 9422/88 $300+000 1988 PergdmonJournalsLtd

NEW HALOMETABOLITES MAURICE

Laboratory

of Orgamc

C. R

FRANSSEN,

Chemistry,

FROM

MAARTEN

Agricultural

CALDARIOM

A. POSTHUMUS

Umverslty,

De Drqen

and

HENK

YCES FUMAGO C.. VAN DER PLAS

5, 6703 BC Wagemngen,

The Netherlands

(Rewed rerrtved 17 Augwt 1987) Key Word Index-Caldarlomyces fumago, Deuteromycetes, ethanol, chloropropanols, chlorohydroxyphenylethanols

halometabohtes,

halogenated

compounds,

dlchloro-

Abstract-The culture medmm of the mould Culdariomycesfumago was extracted with three orgamc solvents and the extracts were analysed by GC-MS Twelve halogen-contaimng metabohtes were found m the dlchloromethane extract Seven of these compounds were sufficiently abundant to allow the identification of the new halometabohtes, 2,2-dichloroethanol, 1,3-dlchloropropanol-2, 1,1,3-tnchloropropanol-2, 1,1,3,3-tetrachloropropanol-2, 2-(3-chloro-4hydroxyphenyl)ethanol and 2-(3,5-dichloro-4-hydroxyphenyl)ethanol m addltlon to the well-known caldarlomycm (2,2-dichloro-lS,3Sdihydroxycyclopentane). Their possible blosynthetlc routes and the lmphcatlons for the reaction mechanism of the Culdanomyces halogenatmg enzyme chloroperoxidase are dlscussed

INTRODUCTION

The occurrence of haloorgamc compounds m hvmg matter is a rapidly growing field of Interest m natural product chemistry Over 700 halometabohtes are known to date The subject has been reviewed m general [l] as well as for mdlvldual classes of orgamsms [2-71. Some of the authors concerned also briefly discuss the blosynthesls of these compounds [2, 4-71 The role of the enzymes responsible for the actual halogenatlon step has, until recently [8], only been mentioned without further comment. This is remarkable, smce the discovery of the caldarlomycm (2,2-dichloro-lS,3Schloroantlbiotic dlhydroxycyclopentane) m the culture medium of the mould Cnldarlomyces fumago [9] directly induced the search for and the detection of the halogenatmg enzyme chloroperoxldase [lo, 111, now one of the most studied halogenatmg blocatalysts. This very versatile enzyme has been used in several laboratories for the synthesis of chlorinated or brominated steroids [12], alkenes [ 13, 147, cyclopropanes [15] as well as chlorinated or brommated derlvatlves of the heterocycles thiazole [ 161, antlpyrme [ 173 and barblturic acid [18-211 In these mvestlgatlons as well as m those from other laboratories [22, 231 conflicting results were presented regarding the reaction mechanism of the enzyme; in particular the stereospeclficlty of the halogenatlon reaction was a matter of debate. Some of the controversy may have been caused by the fact that only a minority of the researchers used cyclopentanedlone, the only known natura1 substrate of the enzyme It would be interesting to know whether there are other natural substrates for chloroperoxidase, and whether this enzyme could halogenate them m a stereoor regiospeclfic manner. Therefore, we wanted to have a more detailed picture of the natural products, formed in the culture medium of C.fumago. When Clutterbuck et al. [9] extracted the culture medmm of C. fumago with ethyl acetate (EtOAc), they not only isolated white crystals of caldariomycm, but also noted that “an oil remained finally from which 1093

nothing crystalhne has been separated, but which contains considerable amounts of Cl m orgamc comblnation” This mduced us to re-examme the culture medium usmg GC-MS. Halogen-contaimng compounds are readily recognized by this technique because of the characteristic 3sCl/37C1 and 79Br/81Br isotope patterns In this paper we report on the extraction and characterization by GC-MS of SIX new chloro-contammg metabolites from the culture medmm of C. fumngo. The blosynthesis of these compounds and their implications for the reaction mechanism and possible stereospeclficlty of chloroperoxldase are discussed. RESULTS AND DISCUSSION

Culdariomyces fumago was grown as described m the Experimental and the filtered culture medmm was extracted successively with petrol 60-80”, CH,CI, and EtOAc, m order to separate the polar from the less polar compounds. The extracts were analysed by GC-MS The petrol fraction was practically devoid of any compounds and was not analysed further. Gas chromatography of the EtOAc fraction and the EtOAc contmuous extract showed 2,3-butanediol, caldariomycm (l), 2-(4hydroxyphenyl)ethanol (9) and an unhalogenated compound of unknown structure* as major constituents. We concentrated our efforts on the dlchloromethane (CH,CI,) extract because of the large number of compounds present therein The CH,CI, extract appeared to contain a considerable number of small, relatively simple unhalogenated molecules as well as 12 chlorinated orgame compounds, one of which was caldarlomycm (1) Five of the 11 new chlorometabolites of C. fumago were only present m very small amounts and could not be

*The mass spectrum of tlus compound contams the followmg peaks. m/z (rel. mt.): 27 (6), 29 (6), 31 (12), 39 (16), 41 (31), 51 (7), 53 (ll), 55 (9), 65 (6), 69 (25), 81 (7), 83 (6), 97 (lOO), 98 (7), 109 (12), 110 (6), 111 (15), 128 (52)

M C R

1094

FRANSFN

et al

Cl,CH-CHzOH

1

2

OH

OH

OH

C12CH-CH-CHCLI

CL2CH-CH-CH,Cl

CH,-CH-CH2Cl 3

5

4

CL

Y Cl

HO&CH,CH,OH

HO

6 analysed further. SIX compounds, however, gave good mass spectra and comparrson wrth hterature data and authentrc samples allowed therr complete rdenttficatron The SIXnew halometabohtes could be dtvlded mto two groups, chloroethanols/propanols and chlorohydroxyphenylethanols (Table 1) The first group conststs of 2,2dtchloroethanol (2), 1,3-dichloropropanol-2 (3) 1,1,3-trtchloropropanol-2 (4) and 1,1,3,3-tetrachloropropanol-2 (5) The structure of 2 was asstgned by compartson of Its mass spectrum with hterature data [24] The structure of 3 was estabhshed by compartson of Its mass spectrum with authentrc material, further support was obtained by the fact that 3 m the CH,Cl, extract coeluted with authentic 3 on GC-analysts The fragmentation patterns of .F5 are very stmtlar the [M]’ peak 1s not vtsrble, but the spectra show charactertsttc peaks at m/z 79 and 113. correspondmg to the a-cleavage products These fragments subsequently lose HCI (peaks at m/z 43 and 77, respecttvely). Compounds 3-5 also showed a good corre-

Table 1 Halometabohtes and related compounds found m the CH,CI, extract of the culture medmm of Ca~dar~omycec fimayo Total ion current (arblWary umts)* 539 I226 2827 9695 3061 1313

Compound no 2 3 4 8 5 1

3111 1285

6

1094 4686

9 7

Compound 2,2-Dlchloroethanol 1,3-Dlchioropropanol-2 1,1,3-Tnchloropropanol-2 2-phenylethanol 1,1,3,3-Tetrachloropropanol-2 Caldarlomycm (2,2-dlchloro-1 3Sdlhydroxycyclopentane) phenylacetlc acid 2-(3-Chloro-4hydroxyphenyljethanol 2-(4-Hydroxyphenyl)ethanol 2-(3,5-Dlchloro-4hydroxyphenyl)ethanol

S-

-2-J'

'

7 latron between their R,s and then A4,, as was found earlier for a serves of halogenated tsopropanols [25] The mass spectrum of 5 IS m good agreement wtth that published previously [25] The second group of halometabohtes consists of 2-(3chloro-4-hydroxyphenyl)ethanol(6) and 2-(3,5-dtchloro4-hydroxyphenyl)ethanoi (7) Their structures were confirmed by comparing then R, and mass spectra wtth those of authentic samples prepared by reaction of 2-(4hydroxyphenyl)ethanol wtth HOC1 These chlorometabohtes have not been described previously However, some analogues are known. 2-Phenylethanol (8, phenetol) and 2-(4-hydroxyphenyl)ethanol(9, tyrosol) are wellknown compounds occurrmg m many moulds, and Murasmn~s palmwxws produces 3-chloro-4-hydroxyphenylacetrc actd [SJ These metabohtes are derived from the shtktmtc actd route, as 1s caldartomycm [26]. Compounds 6 and 7 are probably synthesized vra the shrkrmtc acid pathway, but another route cannot be ruled out. For instance, C’and~Ia sp. and varrous pathogenic fungi produce (9) as the decomposttton product of tyrosme [.5]. We found large amounts of tyrosol m the culture medium of C. fimago (Table l), so thts may well be a precursor for the chlormated tyrosols 6 and 7 To test thts hypothests we incubated tyrosol wtth the pure Caldarromyces halogenatmg enzyme, chloroperoxtdase, and found that tt was readily converted to 2-(3-chloro-4-hydroxyphenyl)ethanol (6) Remarkably, 2-(3,5-dtchloro-4-hydroxyphenyl)ethanol (7) was formed m only small amounts, even after prolonged mcubatton. Smce It IS well-known that tyrosme 1s a substrate for chloroperoxtdase [27], four possrbtlmes for the formation of the chlormecarbon bonds ulttmately leading to 6 and 7 can be advanced. (I) chlormatron of tyrosme residues m protems [28], (11)chlorinatton of free tyrosme, related or not to protein synthesis or degradation, (m) chlormatron of tyrosol and (IV) chlormatron of mter-

/\ o-

CH,CH,OH

-

* Indlcatlon the extract

of the relatwe

concentration

of the compound

m

CH,CH,OH

8

Ho-(pH2cRIoH 9

Halometabolltes from Caldarzomyces

1095

mediates in the shiklmlc acid route, as found for caldarlomycin (1). Since 6 can be formed from tyrosol by chloroperoxldase, as already noted above, all four possible biosynthesis routes may be valid for 6. However m the case of 7 route (iii) is unlikely because 6 proved to be a poor substrate for the enzyme. More information 1s needed before the biosynthesis of 6 and 7 1s known with certainty. The origin of the dichloroethanol and chloropropanols (2-5) 1s even more intngumg. Compounds of this type, are only known from marine organisms. 1,3-Dlchloroted by GC-MS. Until these molecules are fully characterpropanol-2 (3) has been found m Antartic krdl [29], and ized we cannot conclude that chloroperoxidase can cataWoolard et al. [25] detected a large number of haloacereactions in a stereospeclfical manner. tones and haloisopropanols m the CH,Cl, extract of lyse halogenatlon vacuum-dned Asparagopsw taxlformis (Rhodophyta, Bonnemalsoniaceae), including our compound 5. The EXPERIMENTAL halometabolites 2 and 4 have not been found before The General Chloroperoxldase (CPO) was Isolated as described above authors did not study the biosynthesis of these previously 1371 Extra&Ion solvents were dust prior to use. compounds, but since neither ethanol nor acetaldehyde, Growth of mould and lsolatlon of metabohtes C jiumayo propanol-2 or acetone are substrates for chloroperoxlWoron strain CBS 123 26 was obtained from the Centraalbudase, we feel that metabohtes 2-5 must be derived from reau voor Schimmelcultures, Baarn, The Netherlands The larger molecules. Perhaps these larger precursors are molecules like compound 10, from Perrcoma macrosprn- mould was grown on Czapek-Dox medium, conslstmg of 50 g of 1 g of KH,PO,, 0 5 g of KCI, 0 5 g of osa, and cryptosporlopsin (11, from Phdophora asteris, glucose, 2 g of NaNO,, MgSO, 7H,O, 0.02 g of FeSO, 7H,O and 3 3 g of yeast exOctospora carbomgena and many Cryptosporlopsis spetract per 1of deionized H,O The mould was grown for 7 days m cies). These mould metabolites contam the 1,3-dlchloroa 1 I flask and then transferred to a 20 1 flask where it was grown propanol-2 and 1,3-dlchloroacetone moiety, respectively, for 9-10 days. Subsequently the cells were centrifuged and the and possibly these compounds are the precursors of the cell-free supernatant coned m an Amicon hollow-fibre apparcompounds 2-5 or the correspondmg chloroacetone deatus with an H,,P,,-20 cartrlge (M, cutoff= 10000). The 4 5 1 of rivatives* The hypothesis described above 1s speculative clear culture medium thus obtained was dlvlded mto two equal especially since it does not explain the occurrence of tnportions and each portion was extracted with 3 x 5OOml of petrol and tetrachloroisopropanols The blologlcal function of the halometabohtes 1s 6&80”, CH,Cl, and EtOAc, successively The residue was then extracted continuously with EtOAc for 24 hr The extracts were unclear so far but there 1s some evidence that these dned (MgSO,), coned with a rotary evaporator at 30” and compounds play a role in the defence mechanism of the analysed organisms [S]. It 1s interesting to mention that chloroisoCC-MS analysis The extract (2~1) was Injected mto a GC propanols are bacterlcides [31] and that 2-phenylethanol (8), the major constituent of the CH,Cl,-extract of C. equipped with a SIL19 WCOT capdlary column, 26 m x 0 22mm ID, film thickness 0 2 pm The eluted compounds were detected fumago, shows antimicrobial activity [32, 331. by EIMS (70eV, ion source temp 200”). The mltlal column temp One or more halogenatlon steps are involved some(50”) was maintained for 2mm and then raised to 260” at 6”/mm where in the biosynthesis route for the formation of l-7. Blahks Two blank expts were carried out to correct for the The enzyme responsible for these conversions 1s chloropresence of contammatmg halogenated compounds m the peroxldase, an extracellular enzyme produced m relachemicals used (I) CH,Cl, (3 I) and EtOAc (3 1) were evapd to a tively large amounts by C.fumago. Whether this enzyme small vol on a rotary evaporator at 30” and analysed by GCcan act in a stereospecific manner 1s still not established MS A small amount of C,CI, was detected m the CH,Cl, [19, 34-361 concentrate, the EtOAc concentrate was free of contaminants Studying the stereochemistry of the natural substrates (n) Freshly prepared growth medium (4 5 1) was extd and anaand products of the enzyme possibly can throw light on lysed as described above The petrol fraction only showed this matter Five of the halometabohtes found here are solvent peaks The CH,Cl, fraction contained 5-(l-hydroxyachiral, 4 has a chiral central C-atom, but since it was not ethyl)drhydro_2(3H)furanone and some nitrogen-containing possible to isolate this compound its stereochemistry Samples of the EtOAc excould not be studied. The key step m the biosynthesis of compounds of unknown structure tract and the EtOAc-contmuous extract only contained 5-(1caldariomycin (1) is the conversion of cyclopentane-1,3hydroxyethyl@hydro-2(3H)furanone, which was also found m dione into its 2,2-dlchloro denvatlve, again a reaction the CH,Cl, extract These compounds most probably origmate with no need for stereochemlcal control We conclude from the yeast extract m the growth medium. from the structures of the compounds found in this study MS The spectra of the SIX new halometabohtes from C. that chloroperoxidase has no need to be a stereospecdic fumago (2-7) are shown below, only peaks with mtenslties higher chlorinating agent. One should, however, bear m mind than 5% are given that the growth medium of C myces probably contains 2,2-Dtchloroethanol(2) m/z (rel mt.) 31 (100); other important besides the small and volatile halometabohtes larger nonpeaks 29 (3), 43 (4), 48 (3), 49 (4), 79 (3), 83 (2) volatile halogenated compounds which cannot be detec1,3-Dzchloropropanol-2 (3) m/z (rel mt F 27 (6), 43 (42), 49 (7), 57 (9), 79 (lCO), 81 (32)

‘Brewer’s yeast can convert dlchloropropanol-2

[30]

l,l-dlchloroacetone

into

1,1-

1,1,3-Trlchloropropanol-2 (4). m/z (rel. mt ). 27 (lo), 29 (6), 43 (30), 49 (19). 51 (6), 77 (7), 79 (100; calcd for CzH:5C10 m/z =78.9951, found 78 9954), 81 (31), 113 (17, calcd for C,H:SCl,O

1096

M C R

FRANSSEN

m/z = 1 12 9560, found 1129559), 115 (10, Cdkd for C,H:5C137C10 m/z = 114 9530, found 114 9525) 1,1,3,3-Trtrclchloropropanol-2 (5) m/: (rel mt ) 49 (32). 51 (12), 61 (6), 77 (3X), 7X (6), 79 (12). X3 (18. calcd for CH3’Cl,0 m;: =X2 9455, found 82 9457). 85 (13), 1 13 (100, caicd for C,H:SCI,O m/z=l129560, found 1129560). 115 (62), 117 (10) 2-(3-Chloro-4-hydroxqlphrnyl)ethanol(6) m, I (rel In! ) 51 (14), 77 (22), 105 (5), 107 (7), 141 (100, calcd for C,H,3’CIO m,‘= 141 0106. found 141 0092), 142 (IO), 143 (30). 172 (23. cdlcd for C,H,3’CI0, mz= 1720291, found 1720272), 174 (9~ 2-(3,5-DIG hloro-4-hydrouypher~~~~)~~~u~~~ (7) M:: (rel mt ) 31 (7), 51 (7), 75 (17), 111 (ll), 141 (lo), 175 (100, calcd for C,H:‘CIZO m,‘z= 174 9716. found 17497051, 176 (12), 177 (63), 178 (8). 179 (11). 206 (25, calcd for C,Hi’CI,O, m/z = 205 9900, found 205 99 10). 208 ( 15) Ch[~~rmaf~on o/ 2-(4_hydroryphenyl)ethuno[ (I) By chloroperoxldase (CPO) to a soln of 13X mg (1 mmol) of 2-(4hydroxyphenyl)ethanoI m 0 1 M H,PO,/KOH pH 2 7 contaznmg 20 mM of KC1 was added H,O, In eight portions at 4.5 mm mtervala. and 300 jig of chloroperoxtdaae m three porttona at 135 mm Intervals The mix wds stlrred for 1 ddy, the small amount of orange ppt that formed wds dtacarded The mlxt was extracted x 3 with CH2CI,, the extract dried (MgSO,) and the solvent evapd usmg a rotary evaporator The re4ldue conslsted of 2-(3-chloro-4-hydroxyphenyl)ethanol together with some of the 3,5-dlchloro compound, as wds demonstrated by NMR (90 MHz. CDCl,--DMSO-d,) (II) By HOC1 to a similar soln as described under (I) was added very slowly 186 peg (2 5 mmol) of NaOCl The reaction mlxt was stlrred for 6 hr at room temp and then extracted with CH,Cl, The extract was dried (MgSO,), evapd and analysed The product proved to be 2-(3chloro-4.hydroxyphenyl)ethanol together with a small amount of the 3,5-dlc.hloro compound. AF was shown by NMR

9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

A&nowledqemrnrc-We are much indebted to Mr E de Boer and Dr R Wever (Laboratory of Biochemistry, Umverslty of Amsterdam) for provldmg the growth medium of C fumayo, to Prof Dr E M MeIJer, Dr H E Schoemaker dnd Mr W Boesten (DSM) for stlmulatmg discussIons, and to Dr A Fuchs (Department of Phytopathology, Agricultural Umverslty Wagemngen) for crItIcally readmg the manuscript We gratefully &knowledge the linanclal support of Nddmloze Vennootschap DSM, Heerlen, The Netherlands

26 27 28

29

30. 31 REFERENCES Sluda, J F and DeBernardls, J F (1973) Lloydla 36, 107 Turner, W B (1971) Fun@ Metubohtes Academic Press, New York Femcal, W (1975) J Phycol 11, 245 Faulkner, D J (1977) Tetrahedron 33, 1421 Turner, W B and Aldndge, D C (1983) Funyal Metabohtes II Academic Press, New York Faulkner, D J (1984) Nat Prod Rep 1, 251 Engvlld, K C (1986) Phytothemlstry 25, 781 Neldleman, S L and Gelgert, J (1986) B~ohaloqenatron

32 33 34 35 36 37

er ul

Prmcrpie~, Bnsrc Roler and Apphtarrons Ellis Horwood. ChIchester Clutterbuck, Jr, P W, Mukhopadhyay. S L . Oxford, A E and Ralstrlck, H (1940) Blot hem .I 34. 664 Shaw, P D and Hager, L P (1959) J Am Chem SW 81. 1011 Shaw, P D and Hager, L P (1961) .J Blol Chrrn 236, 1626 Neldleman. S L , Dlassl, P A, Junta, B , Palmere. R M nnd Pan, S C (1966) Tetrahedron I,eurrc-. 5337 Gelgert, J, Neidleman, S L, Dahetos, D J and DeWltt, S K (1983) Appl Enwon M~crohw[ 45, 366 Yamada. H , Itoh, N and Izuml, Y (198% J Bull Chem 260. 11962 Geigert. J , Neldleman, S L dnd Dahetos. D J 11983) .I Em/ Chrm 258, 2273 Neldlemdn. S L . Cohen, A I and Dedn, L (1969) BIOfL’(hno/ Hlor!ly I1 * 1?7 Ashley, P L and Griffin B W (1981) Arch B~ochrn~ &ophyc 210, 167 Franssen, M C R and van der Plas. H C (1984) Ret/ 7,al Chrm Puv\-Bu.c 103, 99 Franssen M C R and van der PI&, H C (1987) Rmorq Chem 15, 59 Laane, C, Pronk, W, Franssen, M C R dnd Vcegel. C (1984) Enzr’me M~croh Trchnol 6, 165 Laane, c‘ , Weyland, A and Franssen, M C R (19X6) Enzvmr MIcrob Teechnol 8, 345 Lambelr, A M dnd Dunford. H B (1983) I &01 Chrm 258, 13558 Grtflin, B W (198.1) Blochem B~oph~s Rrs Common 116. 873 Uncertified Mass Spectral Datd (R S Gohlke. ed) (1963) Spectrum no 439 Dow Chemrcal Co, Midland, Mlchlgan Wooldrd, F X, Moore, R E and Roller, P P (1976) Tetrahedron 32, 2843 BeckwIth, J R, Clark. R and Hager. I_ P (1963) .! Bml Chrm 238, 3086 Hager, L P, Morris, D R , Brown, F S and Fberwem. H (1966) J Bw[ Chem 241. 1769 Mltchell Jr, E D. Lormg, R, Carrawag, K L and Cunnmgham, B (1981) Proc O/&I Acad Scr 61. 43, Chem Abv 96, 158187 Kubota, K, Yokoyama, K, Yamamshl, T and Akatsuka, S (1980) &I~ppon Noyel Kayuku Kul+n 54, 727, Chn,1 Ah\ 94, 63937 Sen. H K (1924) Blochem 2 151, 51 Schaefer, G, Rosenthal, W and Bock, K J (1958) Ga? 1095995. Chem Abe 56, 6100h (1983) The Merck Index, 10th Edn (Wmdholz, M , ed ). p 7094 Merck & Co. Rahway. New York Femcal, W (1974)Tetruhedron Letters 4463 Lee, T D, Gelgert, J, Dahetos, D J dnd Hlrano, D S (1983) Bmchern BmphJ \ Ret Commun 110, XX0 Ramakrlshnan. K , Oppenhulren, M E , Saunders, S and Fisher, J (1983) Bwtheml~try 22, 3271 Kollomtsch, J , Marburg, S and PerkIns, L M (1970) J Am Chem Sor 92. 44X9 Hollenberg, P F and Hager, L P (1978) Mefh Enz_~mol 52, 521