sign in sign up
<<
Home , Dihydroxybenzoic acid, Flavonoid, Trihydroxybenzoic acid

OF BY A PENICILLIUMSP.

T. M. THOMASl & J. R L. WALKER

Department of Plant & Plant Microbial Sciences, University of Canterbury, Private Bag 4800, Christchurch, New Zealand

(Received 17August 1993: accepted 9 December 1993)

ABSTRACT

Thomas, T.M. & Walker J.RL. (1993). Metabolism of quercetin by a Penicillium sp. New Zealand Natural Sciences 20:63-70.

The metabolism ofthe flavonol, quercetin, by a Penicillium species isolated from local soil has been investigated. This Penicillium sp. exhibited maximum growth and degradation capability in static cultures at 25°C and at a substrate concentration of2. SmM. In addition, this fungus could utilise kaernpferol, , , dihydroquercetin, and . The products from the initial cleavage of quercetin were 2,4,6- and, probably, 3,4- . 2,4,6-Trihydroxybenzoic acid was metabolized further via pyrogallol. Subsequent ring cleavage of both aromatic intennediates was by ortho fission.

KEYWORDS: Penicillium sp - - quercetin - metabolism ortho fission.

INTRODUCTION acid (THBA, carboxylic acid; from the A-ring). The enzymes involved were studied The flavonol quercetin is widespread in higher extensively (Westlake & Simpson 1961; Westlake et plants, occurring predominantly as glycosylated al.1961; Childetal.1963; Simpsonetal. 1963) and derivatives (Harborne & Simmonds 1964; Harborne it was found that the first step was catalysed by & Williams 1975; Britton 1983). However there quercetinase, a copper containing dioxygenase, have been relatively few studies ofits degradation by which added across carbons 2 and 4 of the rnicro-organisms most work having been done on its heterocyclic ring to produce an ester and carbon rutin and quercitrin (quercetin 3-mono­ monoxide. Subsequently, several other fungi have rhamnoside) since many microbes could not attack been shown to degrade rutin and other flavonol aglycones, but could degrade their glyco­ derivatives via this pathway (Hattori & Nogchi sylated derivatives. For example; (narin­ 1959; Westlake et al. 1961; Westlake 1963, ; Barz genin 7-rhamnoglucoside) was hydrolysed to give 1971). This pathway appeared to be specific for prunin and but the latter was not metab­ and its specificity extended also to the olized further; this inability to cleave flavonoid nature ofthe sugar substituent(s); a specific glycosi­ aglycones was attributed to theirrelative insolubility dase was induced according to the sugar(s) present. (Cieglar et al. 1971). Structures of quercetin glyco­ A different pathway for quercetin breakdown sides are shown in figure 1. was reported for a species ofPseudomonas; the first Micro-organisms were screened for their ability step being hydroxylation at C8 followed by meta­ to grow on flavonoids by Westlake et al (1959, cleavage of the A-ring to yield eventually oxaloac­ Westlake 1963) who found that Aspergillusflavus etate and DHBA (Schultz et al. 1974; Barz & metabolised rutin and quercitrin to yield rutinose, Weltring 1985). A similar A-ring hydroxylation 3,4-dihydroxybenzoic acid (DHBA, protocatechuic was seen also in the degradation oftaxifolin by a soil acid; from the B-ring) and 2,4,6-trihydroxybenzoic Pseudomonad (Jeffery et al. 1972). By contrast Rhizobium loti metabolized quercetin to phloroglu­

1 Current address: School of Biochemistry, La Trobe University, cinol and DHBAby cleavage ofthe heterocyclic ring Bundoora, Victoria 3083, Australia. (Rao et al. 1991). 64 New Zealand Natural Sciences 20 (1993)

OH OH Quercetin R = H HO Quercitrin R =rhamnose Rutin R =rutinose

OH 0

Figure 1. Structures of quercetin glycosides.

Once released from the parent flavonoid the degrading flavonoids. After incubation for several smaller aromatic compounds can be broken down by days in the dark at 26°C, organisms were isolated ortho-(3-oxoadipate pathway) or meta-fission. The using standard plating procedures and pure cultures meta-fission pathway can also metabolize hetero­ were obtained by repeated subculture. Potato dex­ cyclic structures such as flavonoids (Chen & To­ trose agar (PDA) was used for isolation and routine masek 1991). maintenance of micro-organisms. Shake cultures Despite its widespread occurrence in plants and were grown on an orbital shaker at 300rpm and at soils the metabolism of the quercetin aglycone has 26°C. not been studied extensively and was therefore the focus of this present investigation. SELECTION OF ISOLATE Minimal salts medium (lOOmI) containing MATERIALS AND METHODS O. 7mM quercetin was inoculated 'with spore suspen­ sions from the isolated microorganisms and also a ISOLATION AND GROWTH OF FLAVONOID DEGRAD­ range offungi from the Departmental culture collec­ ING MICRO-ORGANISMS tion. Culture flasks (triplicate), either static or Enrichment culture techniques using a mineral shaken, were incubated for six days and sampled salts medium (2 g 1. 1 KH2PO~, 1 g 1. 1 (NH~)2S0~, pH daily to determine the residual concentration of 5.5) containing 0.5mM quercetin as the sole carbon quercetin. The isolate showing maximum rate of source were used to select microbes capable of quercetin degradation was chosen for further study.

Table 1. Products of quercetin degradation by di1ferent micro-organisms.

Products Organism DHPhN mIBA THEA Phloroglucinol

Butyrivibrio sp C3 + + Eubacteria oxidoreductans + Clostridium orbiscindens + + Pseudomonas sp + Rhizobium loti + + Aspergi /Ius jlavus + + Pu/lularia sp. + + Cryptococcus albidus & C.dijJluens +

Ie DHPhA = 3,4-dihydroxyphenylacetic acid, DHP A = 3,4-dihydroxybenzoic acid, THBA = 2,4,6-trihydroxybenzoic acid) T.M. Thomas & IR.L Walker: Metabolism of quercetin 65

FuNGAL REPLACEMENT CULTURE EXPERIMENTS RESULTS Flasks containing basal medium plus 2.5mM quercitrin or rutin were inoculated and incubated for ISOLATION OF QUERCETIN-DEGRADING MICRO­ four days. Mycelium from each flask was filtered on ORGANISMS. Whatman No.114 filter paper, washed with phos­ Growth on quercetin in the enrichment cultures phate buffer (PH 5.5) and finally resuspended in became apparent after three to four days incubation. 50ml phosphate buffer containing 2.5mM querce­ Subsequent examination onPDA plates showed the tin, quercitrin or rutin and incubated at 26°C for majority ofthe organisms present to be fungi but also three days. some yeasts and bacteria. Three fungi (A, B and C) predominated and were isolated into pure culture. A bacterium was also isolated. Cultures of Fusarium IDENTIFICATION OF METABOLIC INTERMEDIATES so/ani, F. oxysporum and Phanerochaete chrysos­ Culture filtrates were extracted with ethyl ace­ porium from the Departmental culture collection tate or n-butanol, the organic layer concentrated in were alsofound to be capable ofgrowth on quercetin. vacuo, and the residue dissolved in 2ml . Static cultures offungal isolate B exhibited the These extracts were then applied to silica gel TLC highest rate of quercetin degradation (49% utili sa­ plates developed in : dioxane: acetic acid tionafter 5 days) so it was chosen for further study. (BzDiAc, 90:24:4) andflavonoids detected by their Colonies of this fungus on PDA plates produced under long wave (360nm) UV light. slow-growing, fluffy white mycelia which eventual­ Similarly, the formation of phenolic intermediate ly produced blue-green spores; it was identified as a compounds was monitored chromatographicallywith species of Penicillium. the use ofspecific chromogenic spmyreagents (Walk­ Subsequent growth experiments revealed that a er & Taylor 1983, Tillett & Walker 1990). substrate concentration of 2.5mM gave maximum Bands of the concentrated methanolic extract quercetin degradation and highest growth rate. Sim­ were also applied to Whatman 3MM paper and ilar experiments with other flavonoid substrates developed in either BzAcW or BuAcW. After found that rutin, quercitrin, luteolin, dihydro­ viewing the dried chromatograms under UV light quercetin and were also readily me­ (360nm), a narrow test strip was sprayed with tabolised by this Penicillium spwhilst was diazotised p-nitroaniline. The remaining unsprayed only slightly degraded and unaffected (Ta­ zones were cut out, eluted with methanol, and ble 2). Structures of these compounds are exempli­ analysed by UV spectroscopy (Walker & Taylor fied in figure 2. 1983, Tillett & Walker 1990). Incubation ofthe quercetin glycosides rutin and Similarly, keto acid intermediates were isolated quercitrin with the Penicillium sp. resulted in an and chromatographed as their 2:4-dinitrophenyl­ accumulation of the aglycone in the culture filtrate; hydrazone (DNPH) derivatives (Walker & Taylor Table 2: Flavonoid substrates utilized by Pencillium sp. 1983, Tillett & Walker 1990). Substrate Approximate percentage degraded UV SPECTROSCOPY (4 days growth) The concentration of quercetin in the culture medium was estimated spectrophotometrically; sam­ ples (O.5ml) of media were added to 4.5ml ethanol Quercetin 30 or methanol and their UV absorption spectra record­ Quercitrin 100 ed with anHP8452ADiodeArray Spectrophotometer. Rutin 85 The characteristic bathochromic shift in alkaline Dihydroquercetin 95 solutions was used to confirm that quercetin was Kaempferol 100 being analysed. Similar procedures were used for Luteolin 95 the estimation and identification ofother flavonoids Fisetin 65 and phenolics. Bathochromic shifts after additions Galangin 20 ofsodium acetate andmethanolicKOHwere record­ Cyanidin 0 ed. 66 New Zealand Natural Sciences 20 (1993) OH OH OH OH OH

HO HO HO

OH 0 OH 0 OH 0 Quercetin Kaempferol Dihydroquercetin () OH OH OH OH

HO ~~0 OH OH 0 0 OH 0 Luteolin OH Fisetin Galangin OH HO

OH 0 Cyanidin

Figure 2.F1avonoid structures. this suggested that breakdown of the quercetin both spectra showed a slight change with sodium might be the rate-limiting step. acetate and a dramatic bathochromic shift after addition of KOH. REPLACEMENT CULTIJRE EXPERIMENTS Quercetin metabolites could not be detected in ACCUMULATION OF KETO ACIDS growing cultures so replacement cultures, without Replacement cultures containing 2.5 roM quer­ added N-source, were employed to facilitate the cetin plus 5roM Na arsenite (an inhibitor of oxida­ accumulation of quercetin degradation products. tive decarbox'Ylation) were found to accumulate 3- These were sampled daily and the culture filtrates oxo-adipic acid which suggests strongly that phe­ analysed for flavonoid and phenolic compounds by nolic ring fission was via the ortho- fission pathway. chromatography and UV spectroscopy. Using this system several phenolic compounds were detected DISCUSSION and chromatography showed that TIffiA or phlor­ glucinol was produced from day 1 to day 3 whilst Growth and degradation of quercetin by micro­ pyrogallol, and a blue fluorescent compound (uni­ organisms, including Penicillium sp. was generally dentified), became obvious after day 3. Similar slow and this may be related to the low solubility of replacement experiments with putative phenolic the substrate. However the rate of degradation intermediates showed that TIffiA was utilised faster increased after several days once mycelial growth than phloroglucinol or pyrogallol. was apparent which suggests that the fungus re­ The band corresponding to the pyrogallol stan­ quired time to adapt and induce the degradative dard also yielded an identical UV spectrum. Neither enzymes needed to produce more readily utilizable spectrum changed when sodium acetate was added metabolites. but after addition of KOH both solutions turned Static cultures supported the best growth and brown and showed similar bathochromic shifts. degradation characteristics for all organisms tested. The phloroglucinol!fHBA spot was difficult to This was in contrast to flavonoid decomposition by resolve but use of TLC and the BuAcW solvent yeasts which required a high level of aeration and a system confirmed that it was the latter. It gave a nutritionally complex medium although glycosides spectrum comparable to that of authentic TIffiA; were degraded in a simple synthetic medium (West- T.M. Thomas & J.R.L Walker: Metabolism of quercetin 67 lake & Spencer 1966). Similarly, Pullularia for­ chromatographically from its Rf value and charac­ mentans only metabolized quercetin when sucrose, teristic colour with diazotized p-nitroaniline. rutinose ornucleic acids were presentin the medium The extra bands which appeared on chromato­ (Noguchi 1963); by contrast our Penicillium sp. grams were separated and analyzed spectroscopical­ degraded quercetin in a minimal salts media. ly. It was thought that the yellow fluorescing band Utilization ofother flavonoid compounds by our was an early cleavage product because its spectnuit Penicillium sp. may reflect the structural specificity was very like that of quercetin and a similar, but of the degradative enzymes; thus kaempferol was slightly different, bathochromic shift was observed totally degraded whilst utilization ofgalangin (no B­ in KOH; this band was not characterised further. ring -OH groups) and quercetin (two B-ring -OH THBA proved more difficult to identify since groups) was comparatively low. From these results only low levels accumulated in the culture medium it seems that thepattem ofB-ring hydroxylation was and it ran very close to phloroglucinol on chromato­ important since it is unlikely that subsequent degra­ grams developed in BzAcW. The compound was dation of the heterocycle cleavage product, thought also unstable, even when stored at 4°C. Phloroglu­ to be DHBA, will be rate-limiting. cinol was also utilized by the quercetin-induced Absence ofthe A-ring C5 hydroxyl group, as in enzyme system whereas pyrogallol was only slowly fisetin, decreased the extent of degradation whilst removed from the medium over four days. the ready utilisation ofluteolin and dihydroquercetin From these experimental results, a tentative suggests that the C3 hydroxyl group or the C2-C3 pathway (Fig. 3) for quercetin break down by Pen­ double bond were not important for recognition or icillium sp. is proposed with the initial cleavage of degradation of these structures. the heterocylic ring yielding THBA, and probably Addition of sugars to the C3 hydroxyl group of DHBAalso. Whilst the former was isolated from the quercetin increased the percentage of substrate de­ culture medium, the latter could not be detected; graded since glycosylation enhanced water solubil­ probably because it is a commmon starting point for ity. However both quercitrin and rutin are also only both ortho- annd meta-ring fission pathways and slightly soluble, so solubility was probably not the was metabolised as fast as it was formed. only reason for the observed increase in catabolism. Conversion of THBA to pyrogallol could pro­ Different sugar substuents may affect degadability ceed via either of the two routes shown. Pyrogallol since it appears that removal of the glycoside moi­ may have been formed in a single step involving the eties usually preceded attack on the flavonoid agly­ loss of CO 2 and the C4 hydroxyl group or via a cone (Westlake et al. 1959). Therefore, since quer­ phloroglucinol intermediate but phloroglucinol could cetin accumulated in culture filtrates ofPenicillium not be identified by UV spectrophotoscopy or chro­ sp. grown on rutin or quercitrin, the first degradative matography; however it was slowly metabolised by steps must have involved removal of the sugars. the quercetin degrading enzyme system. This could lead to accumulation of flavonoid agly­ The initial products of quercetin cleavage are cones in the environment thus creating a significant identical to those reported in earlier studies withA. role for flavonoid-metabolising microbes. jIavus (Westlake et al. 1959; Westlake 1963) but Quercitrin (3-rhamnoside) was degraded more these workers did not investigate their subsequent rapidly than rutin (3-rutinoside); both compounds metabolism. Similar metabolites were reported also contain rhamnose but rutin also has a glucose group in studies with a Pullularia species, Cryptococcus attached to the C3 hydroxyl and the Penicillium sp. albidusand C. d@uens(Westlake&Spencer 1966). glucosidase may be less active than its rhamnosi­ Thus it appears that this flavonol-cleavage pathway dase. However, in A.jlavus, rutinose was cleaved is common in fungi. from rutin and was not degraded further (Westlake et al. 1959). METABOLISM OF 2,4,6-TRIHYDROXYBENZOIC ACID Several studies have provided evidence for THBA METABOLIC INTERMEDIATES breakdown by microorganisms. SchinkandPfennig Putative intermediates in the quercetin break­ (1982) found that the strict anaerobe Pelobacter down pathway were identified in several ways. acidiga/lici metabolised THBA, pyrogallol, phloro- . Pyrogallol was identified from its UV spectra and glucinol and (3,4,5-trihydroxybenzoic 68 New Zealand Natural Sciences 20 (1993)

OH OH

HO Heterocyclic ring cleavage products

OH / ~OH HO \~OH OH Quercetin + , o , , THBA ~DHBA , , / COOH , , OH

I O~COOH /",,/" ? :?I ~COOH I ~ )!,' • 3-Oxo-adipic ~ ____? __ HO~OH acid r®l HO~OH OH OH Pyrogallol Phloroglucinol

Figure 3.Tentative pathway for the metabolism of quercetin by a Penicillium sp. acid) via phloroglucinol. THEA and DHBA were Penicillium patulum was found to produce large produced from by the bacteri­ quantities ofpyrogallol (Tanenbaum & Bassett 1958) um Bradyrhizobium japonicum (Hopper & Ma­ whilst another Penicillium sp. accumulated pyro­ hadevan 1991) and which converted THEA to phlo­ gallol when grown on medium (Yoshi­ roglucinol; the latter being metabolised via resorcin­ da et al. 1982). ol to hydroxy-quinol and thence by ortho-fission to Anaerobic bacteria, such as P. acidigallici, 3 -oxo-adipic acid. This pathway for phloroglucinol Pelobacter massiliensis (Schnell et al. 1991) and metabolism was found also inPenicillium simplicis­ Eubacterium oxidoreducens (Krumholtz & Bryant simum (patel et al. 1990) whereas Azotobacter 1988) appear to convert p)Togallol to acetate via vinelandii converted phloroglucinol to resorcinol phloroglucinol. The pyrogallol-phloroglucinol and thence to pyrogallol (Groseclose & Ribbons isomerase of E. oxidoreducens was characterized 1981). However, neither resorcinol nor hydro;...')'­ and a conversion mechanism proposed (Krumholtz quinol were detected in our Penicillium sp. culture & Bryant 1988). Thus it is possible that this, or a filtrates. Metabolism of phloroglucinol via pyro­ similar phloroglucinol to pyrogallol interconversion, gallol, without detectable resorcinol intermediates, may operate in Penicillium sp; however, this would was reported in Fusarium solani (Walker & Taylor depend on favourable thermodynamics and enzyme 1983). kinetics for the reverse reaction. Groseclose and Ribbons (1981) and Walker and PYROGALLOL METABOLISM AND MODE OF RING Taylor (1983) demonstrated that pyrogallol could be FISSION cleaved by meta fission. However, in the present The utilization of pyrogallol by microbes has study, only evidence for ortho fission was encoun­ not been studied in detail. Mills et al. (1981) tered; thus it was concluded that pyrogallol and compared growth of yeasts on aromatic substrates DHBA both underwent ortho fission. It is possible and found that none of the yeasts tested could grow that, in our Penicillium sp., pyrogallol ring cleavage on pyrogallol. A non-pigmented mutant strain of could be a rate-limiting step for quercetin break- T.M. Thomas & J.R.L Walker: Metabolism of quercetin 69 down ifhigh concentrations of pyrogallol inhibited dation of rutin. Nature 184: 1145-1146. the enzymes involved early on in quercetin break­ Hopper, W. & Mahadevan, A (1991). Utilization down. ofcatechin and its metabolites by Bradyrhizobi­ umjaponicum. AppliedMicrobiologyandBio­ ACKOWLEDGEMENTS technology 35: 411-415. Krumholz,L.R. & Bryant, M.P. (1988). Character­ The authors wish to thank Dr A L. 1 Cole for ization ofthe pyrogallol-phloroglucinol isomer­ mycological advice, Mrs 1 Healyfor technical assis­ ase ofEubacterium oxidoreducens. Journal of tance and the University of Canterbury for provision Bacteriology 170: 2472-2479. of research facilities. Mills, C., Child, 11 & Spencer, IF.T. (1971). Utilization of aromatic compounds by yeasts. REFERENCES Antonie van Leeuwenhoek 37: 281-287. Noguchi, N. (1963). Degradation offlavonols by Barz, W. (1971). Metabolism of aromatic com­ Pullularia fermentans var. candida. Botanical pounds by Fusarium oxysporum Schlecht. Ar­ Magazine (Tokyo) 76: 191-198. chiv fur Mikrobiologie 78: 341. Patel, T.R., Hameed, N. & Martin, A.M. (1990). Barz, W. & Weltring, K.M. (1985) Biodegradation Initial steps of phloroglucinol metabolism in of aromatic extractives of . In Biosynthe­ Penicillium simplicissimum. Archives ofMi­ sis and Biodegradation of Wood Components crobiology 153: 438-443. (ed. T. Higuchi), pp607-666. Academic Press, Rao, IR., Sharma, N.D., Hamilton, J.T.G., Boyd, Florida. D.R. & Cooper, IE. (1991). Biotransformation Britton, G. (1983). The Biochemistry of Natural of the pentahydroxy flavone quercetin by Rhi­ Pigments. pp 102-113, 124-127. Cambridge zobium loti and Bradyrhizobium Strains. Ap­ University Press, Cambridge. pliedandEnvironmentalMicrobiology 57: 1563- Chen, C.M. & Tomasek, P.R. (1991). 3,4-Dihy­ 1565. droxyxanthone dioxygenase fromArthrobacter Schink, B. & Pfennig, N. (1982). Formation of Sp Strain Gfb100. Applied and Environmental trihydroxybenzenes by Pelobacter acidigallica Microbiology 57: 2217-2222. gen. nov. sp. nov., anew strictly anaerobic, non­ Child,II,Simpson,F.I&Westlake,D.W.S. (1963). sporeforming bacterium. Archives ofMicrobi­ Degradation of rutin by Aspergillus jlavus - ology 113: 195-201. production, purification and characterization of Schultz,E.,Engle,F.E.&Wood,J.M. (1974). New an esterase. Canadian Journal ofMicrobiology oxygenases in the degradation of and 9: 653-664. by Pseudomonas putida. Biochem­ Ciegler, A, Lindenfelser, L.A. & Nelson, G.E.N. istry 13: 1768-1776. (1971). Microbial transformationofflavonoids. Simpson, F.I, Narasimhachari, N. & Westlake, Applied Microbiology 22: 974-979. D.W.S. (1963). Degradation of Rutin by As­ Groseclose, E.E. & Ribbons, D.W. (1981). Metab­ pergi llus jlavus: The carbon monoxide produc­ olism of resorcinylic compounds by bacteria: ing system. Canadian Journal ofMicrobiology new pathway for resorcinol catabolism inAzo­ 9: 15-25. tobacter vinelandii. Journal of Bacteriology Tillett,R. & Walker,IR.L. (1990). Metabolism of 146: 460-466. by a Penicillium sp. Archives of Harborne, J.B. & Simmonds, N.W. (1964). The Microbiology 154;206-208. Natural Distribution ofthe Phenolic Aglycones. Walker, J.RL. & Taylor, B.G. (1983). Metabolism In Biochemistry ofPlant Phenolics (ed. Har­ ofphloroglucinol byFusarium solani. Archives borne), pp 77-128. Academic Press, New York. ofMicrobiology 134: 123-126. Harborne, IB. & Williams, C.A. (1975). Flavone Westlake, D. W.S. (1963). Microbiological degra­ and flavonoid glycosides. In The Flavonoids dation of quercitrin. Canadian Journal ofMi­ (ed. Harboume, Mabry, Mabry), pp 376-441. crobiology 9: 211-220. Chapman and Hall, London. Westlake, D.W.S., Roxburgh, J.M. & Talbot, G.' Hattori, S. & Noguchi, I. (1959). Microbial degra- (1961). Microbial production ofcarbonmonox- 70 New Zealand Natural Sciences 20 (1993)

ide from flavonoids. Nature 189: 510-511. Westlake, D.W.S., Talbot, G., Blakley, E.R & Westlake, D.W.S. & Simpson, F.l (1961). Degra­ Simpson, F.J. (1959). Microbial decomposi­ dation of rutin by Aspergillus flavus. Factors tion of rutin. Canadian Journal ofMicro biol­ Affecting Production of the Enzyme ~ystem. ogy 5: 621-629. Canadian Journal afMicrobiology 7: 33-44. Yoshida, H., Tani, Y. & Yamada, H. (1982). Westlake, D.W.S. & Spencer, IF.T. (1966») Utili­ Isolation and identification of a pyrogallol pro­ zation of flavonoid compounds by yedst and ducing bacterium from soil. Agricultural and yeast-like fungi. Canadian Journal ofMicrobi­ Biological Chemistry 46: 2539-2546. ology 12: 165-174.