50 J.C.S. Perkin IStudies in Relation to Biosynthesis. Part X1W1 Incorporation ofcyc/o-L-Tryptophyl-L-proline into Brevianamide ABy John Baldas, Arthur J. Birch," and Richard A. Russell, Research School of Chemistry, AustralianNational University, P.O. Box 4, Canberra ACT 2600, AustraliaThe culture conditions and method of isolation of brevianarnide A (1 ) from Penici//ium brevicompactum have beenimproved. Use of SH and 14C labelling has shown that cyclo-L-tryptophyl-L-proline (5) is incorporated efficientlyinto brevianamide A (1) and a biosynthetic sequence is proposed.BREVIANAMIDE A (l), isolated as the major componentof a related group of pigments from Peizicillium brevicom-pactzlm,2 was the first characterised of a growing numberof fungal dioxopiperazines which may be derived fromtryptophan, proline, and one or more units of mevalonicacid.Steyn4 has isolated the related austamide (2)and congeners from Aspergillus ustus, including the pre-(1, I-dimethylallyl) tryptophyl-L-proline (3). Lanosulinazine nitrogen atom, either in an indole or in an indoxylprecursor. A possible process for the latter is shown (A).f7q,,,$ 0 N 0H( 2 )dicted biosynthetic precursor of (1) * and (2), cycZo-~-2-k ( 1 1 i 7 )SCHEME(4), recently isolated 6 from Penicillium lanosum, differsfrom the brevianamides and echinulin (7) in lacking anisoprene unit attached in the reversed manner to theposition of the indole; both structures (2) and (4)have a C, unit (or units) attached in the normal mannerto nitrogen. The reversed unit could be attachedtransferred in the brevianamides from the dioxopiper-1 Part XLV, A.J. Birch and D. J. Thompson, Austral. J .Chern., 1972, 25, 2731.2 A. J. Birch and J. J. Wright, Tetrahedron, 1970, 26, 2329.3 A. J. Birch and J. J. Wright, Chem. Comm., 1969, 644.An alternative possibility is ring formation leading tos t r ~ t u r e (2) by s~mdtaneous addition to carbon and d N 2 N~ 0 H+N HAamp;H directly through a cation or might be biogenetically( A )P. S. Steyn, Tetrahedron, 1973, 29, 107.D. T. Dix, J . Martin, and C. E. Moppett, J.C.S. Chenz.Comm., 1972, 1168.1974 51nitrogen in a carbene-type reaction, such as that postu-lated for the introduction of a terpene unit into a non-phenolic benzene ring.Biogenetic considerations and physical methods werecombined to elucidate the structure of (l).293 In theinitial work the incorporation of x4C-labelled DL-tryPtO-phan, DL-mevalonate, and L-proline into (1) was low(0.6, 0.003, and 0-09, respectively) probably owinglargely to the instability and the very low yield of (1)produced.In the present work it has been found thatimproved culture conditions, including the partial re-placement of the culture medium by sterile water priorto precursor feeding, give more efficient incorporations.Also, owing to a reduced growing period, brevianamide A(1) can be isolated from the mycelium instead of fromthe medium.The isolation of cyclo-L-tryptophyl-L-proline (5) fromthe culture medium of Penicillium brevicompactum to-gether with the formation of (3) by Aspergillus ustussuggests that the sequence leading to brevianamide A (1)may be as shown in the Scheme.The postulated pre-cursor (3) has, however, not so far been detected incultures of P. brevicompactum.It has been shown 8 that -labelled cyclo-L-trypto-phyl-L-alanine (6) is incorporated efficiently intoechinulin (7) by Aspergillus amstelodami and that a cell-free extract will convert a mixture of 3,3-dimethylallylpyrophosphate and (6) into a derivative with the singleisoprene substituent 9~10 on the pyrrole ring.Table 1 shows the incorporation of ~~-methylene-l~C-tryptophan, L-~-~H proline, DL- 2-I4C mevalonic acidTABLE 1Incorporations into brevianamide A (1)Activity fed Total SpecificPrecursor t m incorporation incorporationDL- methyZene-14C- 100 5.1~-5-3HProline 100 0.40DL-2-l4C- 75 0.16Tryptophan 100 17-6Mevalonic acidlactone aTryptophyl-L- 12.6 14.0prolinethe L-isomer only.cyclo-~-methylene-~*C 12.9 1.8a Incorporations are based on the expected utilisation oflact one, and cycZo-~-methylene-~~C J t ryp t ophyl-L-proline(5) into brevianamide A (1).The specific molar in-corporations were determined by feeding precursors ofknown specific molar activity and recrystallising thebrevianamide A produced to constant activity withoutthe addition of inactive material. The results showthat the total activity recovered as brevianamide A isconsiderably less than the actual total incorporation.This is undoubtedly due to the instability of brevianamideA during the fermentation and isolation procedures.To establish that (5) is incorporated as a unit intobrevianamide A without any prior hydrolysis into the' A.J. Birch and R. A. Russell, Tetrahedron, 1972, 23, 2999.A. J. Birch, J . Agric. and Food Chem., 1971, 19, 1088.amino-acids, followed by recombination, cyclo-~-mdhyl-ev~e-~~C tryptophyl-~-5-~Hproline was prepared. Feed-ing of the doubly labelled material (3H : 14C ratio 3.67 : 1 ;specific activity 306 x lo3 and 83.5 x lo3 disint. min-lmg-l for 3H and 14C) gave (Table 2) brevianamide A withTABLE 2Incorporation of C~CZO-L- methyZene-14C tryptophyl-L-5-3Hproline into brevianamide ABrevianamide APrecursor3H : 14C Ratio 3H:"CRatioTotal incorporatiamp;3-67 : 1 (105 mg) 3.82 : 1 3.23-67 : 1 (106 mg) 3.74: 1 3.63H : 14C ratios of 3.82 : 1 and 3.74 : 1 and total incorpora-tions of 3.2 and 3.6, respectively.The results showthat cyclo-L-tryptophyl-L-proline (5) is incorporatedintact into brevianamide A (1). Furthermore, the highspecific incorporation of (5) (14.0) confirms the routeshown in the Scheme. The introduction of an isopreneunit into the indole ring of (5) followed by furtherelaboration of (3) to brevianamide A (1) must represent amajor (if not the only) biosynthetic sequence. Thedetails of the final stages are still unknown.EXPERIMENTALM.p.s were determined on a Kofler hot-stage apparatus.Analytical and preparative t.1.c.was carried out with MerckKieselgel GF,,*. Radioactive samples were counted on aBeckman LS- 160 liquid scintillation counter (internal stand-ardisation with 3H and 14C standards).Culture and Isolation of Labelled Brevianavnide A (1) .-Astrain was selected of Penicillium brevicompactunz Dierckx(University of Manchester Acc. 382) which when culturedon corn-steep liquor-glucose slopes produced dark greenspores. A medium consisting of glucose (20 g), corn-steepliquor (11 g), and distilled water (500 ml) in 2 1 penicillinflasks was inoculated with a heavy spore suspension of themould and the culture was grown in the dark at 25".Growth was rapid in comparison with cultures grown onCzapec-Dox medium. After 2 days the surface of themedium was covered with thick white mycelium andsporulation had just begun.At this stage ca. 400 ml of themedium was removed and replaced with an equal volumeof sterile distilled water containing the labelled precursor.Alternatively, the labelled precursor dissolved in the mini-mum amount of ethanol was added to the reflooded culture.Concentrations of ethanol greater than 10 ml per 1 of mediuminhibited spore development and subsequently pigmentproduction. By the third day of growth the mycelial matwas completely covered with dark green spores and thecultures were harvested after 7 days. In a typical feedingexperiment the mycelium from five cultures was filtered offand the medium discarded. Care was taken to excludelight in this and subsequent operations in order to minimisethe decomposition of brevianamide A.The mycelium wasdried under reduced pressure (P,O,) and powdered beforebeing stirred with light petroleum (b.p. 60-80"; 600 ml)for 4 h to remove lipids. The light petroleum was removedG. P. Slater, J. C. McDonald, and R. Nakashima, Bio-chemistry, 1970, 9, 2886.9 C. M. Allen, jun., Biochemistry, 1972, 11, 2154.l o C. M. Allen, jun., J . Amer. Chem. Soc., 1973, 95, 238652by filtration and the dark green powder was stirred over-night with methanol (400 ml). The deep yellow solutionwas filtered and evaporated to dryness under reducedpressure. The residue was extracted with ethyl acetate(600 ml) and the resulting yellow solution was washed withsaturated aqueous sodium carbonate (3 x 100 ml) andbrine (1 x 100 ml), dried (MgSO,), and evaporated to asmall volume.Brevianamide A was isolated by preparativet.1.c. on silica (10 ethanol-ether as solvent) and recrystal-lised to constant activity from chloroform ; m.p. 220-250"(sublimation) lit.,2 220-250" (sublimation). In experi-ments where total recovered activity was required inactivebrevianamide A (10-20 mg) was added before crystallisa-tion.Preparation of Lubelled cyclo-L-Tyyptophyl-L-prolines (5).--~ycZu-~- methyZene-1QCTryptophyl-~-5-~Hproline was pre-pared from benzyloxycarbonyl-~-methyZene-~~Ctrypto-phan 11 and ~-5-~Hproline benzyl ester hydrochloride l2by the method described for the unlabelled c~mpound.~The doubly labelled compound was recrystallised to con-stant 3H : 14C ratio (3.67 : 1; specific activities 306 X lo3and 83.5 x lo3 disint.min-l mg-l, respectively) fromethanol ; m.p. 173-1 75' (lit. ,' 17 3-1 75"). cyclo-L- methyl-J.C.S. Perkin Iene-14CTryptophyl-~-proline was similarly prepared frombenzyloxycarbonyl-~-methylene-~~Ctryptophan and in-active L-proline (specific activity 278 x 103 disint. min-1mg-1).Calculation of 3H : 14C Ratios.-The 3H : 14C ratios fordoubly labelled brevianamide A and doubly labelled com-pound (5) were determined by making corrections for thenon-linear relationship between the counting efficiency fortritium, and to a lesser extent for 14C, with the quenchingfactor. A correction was also made for the varying spill-over of 14C counts into the 3H channel with quenching factor.A direct comparison between the doubly labelled (5) fedand the brevianamide A produced can be made. When asample of 3H,W brevianamide A was compared directlywith a sample of 3H,14C-(5) made up to the same quenchingfactor with inactive brevianamide A, the 3 H : 14C ratioswere found to agree to within 3.We thank Mrs. M. Anderson for the culture work.3/1213 Received, 11th June, 1973111 E. I. Smith, J . Biol. Chem., 1948, 175, 39.12 R. E. Neuman and E. L. Smith, J. Biol. Chem., 1961,195,97
展开▼
