Biotechnology Letters21: 763–769, 1999.© 1999Kluwer Academic Publishers. Printed in the Netherlands.

763

Production of aryl metabolites in solid-state fermentations of the white-rotfungus Bjerkandera adusta

Carmen Lapadatescu1 & Pascal Bonnarme2,∗1Laboratoire de Recherches Sur les Arômes, Institut National de la Recherche Agronomique, 17 rue Sully, 21034Dijon cedex, France2Laboratoire de G´enie et Microbiologie des Proc´edes Alimentaires, CBAI, Institut National de la RechercheAgronomique, 78850 Thiverval-Grignon, France∗Author for correspondence (Fax: 33 (0)1 30 81 55 97; E-mail: bonnarme@platon.grignon.inra.fr)

Received 3 June 1999; Revisions requested 28 June 1999; Revisions received 15 July 1999; Accepted 16 July 1999

Key words:aryl metabolites,Bjerkandera adusta, Perlite, solid-state fermentation, wheat bran

Abstract

Bjerkandera adustaproduced aromatic compounds such as benzaldehyde (bitter almond aroma), benzyl alcoholand benzoic acid from L-phenylalanine (3 g kg−1). Two supports for the fungus, wheat bran (organic support)and Perlite (mineral support), gave optimal production with water contents of 66% and 60%, respectively. Benzylalcohol (4.53 g kg−1) and benzaldehyde (1.56 g kg−1) were produced after 4 days on wheat bran respectively with20 and 30 g L-phenylalanine kg−1. Aryl alcohol oxidase activity, which oxidises benzyl alcohol to benzaldehyde,was only detected when the fungus was grown on wheat bran, the support which promotes the most benzaldehydeproduction. Results are compared with those obtained in submerged liquid cultures.

Introduction

Among aroma producers, white-rot fungi are probablythe most interesting microorganisms (Janssenset al.1992, De Jonget al.1994). These fungi employ a va-riety of extracellular enzymes (lignin and manganeseperoxidases, aryl alcool oxidase) involved in the pro-duction of a wide variety of aromatic compounds (Tien& Kirk 1983, Glenn & Gold 1985, Muheimet al.1990, Galloiset al.1990). The aryl alcohol oxidase isof particular interest since it catalyses the oxidation ofvarious aryl alcohols (benzyl, anis, veratryl alcohols)to aldehydes of valuable interest (Muheimet al.1990).

L-Phenylalanine is the precursor used to pro-duce benzaldehyde and benzyl alcohol inBjerkan-dera adusta, Ischnoderma benzoinum, Dichomitussqualens(Lapadatescuet al. 1997) andPolyporus tu-beraster(Kawabe & Morita 1994) in agitated liquidcultures. Solid-state fermentation (SSF) can offer analternative to liquid fermentation for the production ofconcentrated solutions of aroma compounds (Christenet al.1997). Filamentous fungi such as white-rot fungi

are well suited for SSF since these conditions are simi-lar to their natural habitat (Hesseltine 1977). However,few works have reported the production of bioflavoursby SSF (Gervais & Sarrette 1990, Christenet al.1994,1997).

In this work, SSF is explored as a possible al-ternative culture technique for the production of arylmetabolites byBjerkandera adusta.Results are com-pared with those obtained with submerged liquidfermentation.

Materials and methods

Maximum water content measurement

For each solid support, the maximum water content(WCmax) was calculated as follows:

WCmax(%) = (Qw/Wm +Qw)× 100,

Qw = weight of water (kg) retained in the solidsupport;Wm = solid support weight (kg).

764

Micro-organism and culture conditions

Precultures ofBjerkandera adusta(CBS 595.78) werecarried out as previously described (Lapadatescuet al.1997). After 10 days, the fungal pellets were collected,homogenized (Ultraturrax T25, IKA Labortechnik,Staufen, Germany) and used as the inoculum. Perlite(Fertil Boulogne-Billancourt, France) or wheat bran(Moulins Ph. Cochard, Cuisery, France) was usedas solid support for SSF. Wheat bran contained (perdry matter) about 14.9% protein, 6.1% fat, 30.6%starch, 7.5% free sugars, 20.2% pentosans, 7.5%cellulose, 3.1% lignin, 5.6% uronic acids and 4.5%ash. Perlite (10 g) or wheat bran (15 g) were placedin 500-ml flasks and autoclaved at 121◦C (20 min).The following culture medium: KH2PO4 (1 g l−1);MgSO4 ·7H2O (1 g l−1); CuSO4 ·5H2O (0.05 g l−1);yeast extract (2.5 g l−1); lecithin (50 g l−1); L-phenylalanine (15 g l−1) was added to the supports.Various volumes of the culture medium were addedper flask to obtain the desired initial water content.50 µl of the inoculum were added to inoculate eachflask. When the precursor concentration was studied,different L-phenylalanine amounts were added to theculture medium in order to obtain a final concentrationof 10, 20 and 30 g kg−1 support. In these cases, 66%water content for wheat bran and 60% water contentfor perlite support were employed. In submerged liq-uid fermentation, inoculum and cultures were carriedout as previously described (Lapadatescuet al.1997).

HPLC quantitative analysis of aromatic metabolites

After various incubation periods, 110 ml of waterwere added per flask. The flasks were subsequentlyshaken at 100 rpm during 30 min in order to achievehomogeneous suspensions. Then, the liquid mediumwas separated from the substrate and used for aro-matic metabolites quantification. In order to recoverthe metabolites trapped in the support, 110 ml ofmethanol was added and the previous experiment wasrepeated. Final aryl metabolites concentrations takesboth methanol and water soluble fractions into ac-count. Benzaldehyde, benzyl alcohol, benzoic acidand L-phenylalanine were quantified by high perfor-mance liquid chromatography (HPLC) analyses, aspreviously described (Lapadatescuet al.1999).

Enzyme assays

The enzymatic activities were determinated extracel-lularly at 25◦C in the water soluble fraction once

filtered and diluted. Lignin peroxidase (LiP), man-ganese peroxidase (MnP), laccase and aryl alcoholoxidase (AAO) activities were measured as previouslydescribed (Tien & Kirk 1983, Paszczynskiet al.1988,Gutiérrezet al.1994). When detected, activities wereexpressed in U kg−1 support (1 U= 1µmol min−1).

Data analysis

All data were obtained from triplicate assays whichwere repeated at least once. The reported values aremeans " standard deviation.

Results and discussion

Effect of various water contents of the supports onaryl metabolites production byB. adusta

Wheat bran and Perlite were used as solid supports.Their physical properties are given in Table 1. Forboth supports, L-phenylalanine consumption rate in-creased when the initial water content was enhanced(Figure 1A). The highest precursor consumption rateswere 4.08 g kg−1 support day and 2.26 g kg−1 supportday with water content of 66% (wheat bran) and 60%(Perlite), respectively.

Increasing initial water content led to an enhance-ment of the production of aryl metabolites and AAOactivity (Figures 1B–E). Best results were obtainedwith the organic support. Benzyl alcohol was the ma-jor metabolite produced byB. adusta.Benzyl alcoholconcentration reached 3.75 g kg−1 support after 4 dayswith 66% water content on wheat bran. On Perlite,2 g benzyl alcohol kg−1 support was produced after5 days with 60% water content (Figure 1B). Impor-tant amounts of benzaldehyde were biosynthesized onwheat bran, maximum concentration (1.56 g kg−1)being obtained with 66% water content after 4 daysof incubation (Figure 1D). Fairly low amounts of ben-zaldehyde (≤ 0.14 g kg−1 support) were produced onPerlite (Figure 1D). Poor productions of benzoic acidwere obtained on both supports (Figure 1C).

Among the assayed enzymatic activities (LiP,MnP, AAO and laccase), only AAO was detected inB. adustagrown on wheat bran. AAO activity reached1400 U kg−1 support after 5 days of cultivation with66% water content (Figure 1E). The onset of AAO ac-tivity was associated with benzaldehyde biosynthesis(Figure 1D, E). The differences in metabolite produc-tion levels between both supports can be attributed to

765

Table 1. Physical properties of wheat bran and Perlite supports.

Support Support Characteristics Diameter of kg water/ Maximum water

origin particles (mm) kg support content (%)

Wheat bran organic Fine chips 0.7 3.81 79

(plant)

Perlite mineral Gray-white 4.5 1.08 65

(volcanic) siliceous material

Fig. 1. Effect of water content on L-phenylalanine consumption (A), benzyl alcohol (B), benzoic acid (C), benzaldehyde (D) and aryl alcoholoxidase activity (E) production byB. adustacultivated on wheat bran ( 40%,� 50%,N 66% water content) or Perlite (# 40%,� 50%,M60% water content).

766

the basic support structure: wheat bran acts simulta-neously as a support and a substrate for growth, whilePerlite is only an inert support (Durandet al.1996).

Effect of L-phenylalanine initial concentration onaryl metabolites production

As a result of the above study, the initial wa-ter content was 66% for cultures grown on wheatbran and 60% when grown on Perlite. Enhancingthe precursor concentration led an increase in L-phenylalanine consumption rate. It reached 3.9 g L-phenylalanine kg−1 support day on wheat bran and3.5 g L-phenylalanine kg−1 support day on Perlite for30 g kg−1 L-phenylalanine (Table 2).

Benzyl alcohol was the major aryl metabolite accu-mulated byB. adustafor both supports (Table 2). Themaximum benzyl alcohol production was 4.53 g kg−1

support after 4 days on wheat bran and 1.13 g kg−1

support after 6 days on Perlite respectively with 30and 20 g L-phenylalanine kg−1 support. An increase inthe initial precursor concentration, reduced benzalde-hyde production (Table 2). Maximum benzaldehydeproduction (1.56 g kg−1 support) was obtained onwheat bran support with 30 g precursor kg−1 support(Table 2). When Perlite was employed, the maxi-mum benzaldehyde production was only 0.14 g kg−1

with 30 g L-phenylalanine kg−1 initial concentration(Table 2). While maximum benzoic acid concentra-tion was 0.29 g kg−1 after 4 days of cultivation onwheat bran, comparable benzoic acid concentrations(0.09 g kg−1 support) were obtained when the funguswas cultivated on Perlite (Table 2). The highest bio-conversion yields were 52% on wheat bran and 31%on Perlite (Table 2).

As previously observed (Figure 1), aryl alco-hol oxidase activity was only when wheat bran, theonly support which allowed important benzaldehydebiosynthesis, was employed. After 4 days of culti-vation, the highest aryl alcohol oxidase activity was1260 U kg−1 support for an initial L-phenylalanineconcentration of 30 g kg−1 support. This also corre-sponds to the maximum production of benzaldehyde(Table 2). Recent work done by Rocheet al. (1994)has reported that an increase in the initial nitrogen con-centration enhanced the biosynthesis of the enzymeα-L-arabinofuranosidase inThermoascus aurantiacuscultivated on sugar beet pulp. SSF is able to minimizecatabolic repression of enzyme formation (Rocheet al.1994) and consequently enzyme biosynthesis could be

promoted. Therefore, a similar effect withB. adustacan be expected when cultivated in solid-state culture.

Comparison of metabolites production in solid statecultures and in liquid submerged cultures

These results show that the production of aryl metabo-lites is significantly increased and accelerated in SSFwhen an organic support (wheat bran) is used ascompared with submerged liquid fermentation (SLF)(Lapadatescuet al. 1997) (Table 3). Using SSF al-lowed a 10-fold increase in benzyl alcohol produc-tion (3.75 g kg−1 support) while the productivity(0.94 g kg−1 wheat bran.day) was increased 27.5-fold as compared with SLF. Benzaldehyde maximumproduction increased 2.66-fold (1.56 g kg−1 sup-port) in SSF and benzaldehyde productivity 5.34-fold(0.39 g kg−1 d−1) as compared to SLF. Moreover,benzoic acid concentration and productivity were in-creased respectively 3.5-fold and 8-fold in SSF. Thebioconversion yields are comparable in SSF (52%)and SLF (56%). The time to reach maximum metabo-lite production was considerably reduced in SSF(4 days of incubation) as compared to SLF (8 days).Moreover, AAO activity was 47-fold higher in SSF(1400 U kg−1 support) than in SLF (30 U Kg−1) (La-padatescuet al. 1997). This can partly explain thesuperiority of SSF over SLF at producing benzalde-hyde.

Conclusion

These results have demonstrated the importance ofthe support origin, water content and precursor (L-phenylalanine) initial concentration on aryl metabo-lites biosynthesis byB. adustawhen cultivated insolid-state cultures. SSF has revealed a great poten-tial to produce benzaldehyde and benzyl alcohol, sincemore important amounts of these metabolites can beproduced over a shorter period of time than by SLF.

Acknowledgements

The authors are grateful to A. Durand at the Lab-oratoire de Recherches Sur les Arômes for valuablediscussions and technical advices on SSF. CL isgrateful to INRA (Institut National de la RechercheAgronomique), DRI (Direction des Relations Interna-tionales), for a Ph.D. scholarship.

76

7

Table 2. Effect of L-phenylalanine concentration on aryl metabolites production byBjerkandera adustawhen cultivated on wheat bran andPerlite.

Support Time3 L-phenylalanine Benzaldehyde Benzyl alcohol Benzoic acid Yield AAO(2)

(days) Initial Residual production production production (%)1 activity

(g kg−1 support) (g kg−1 support) (g kg−1 support) (g kg−1 support) (g kg−1 support) (U kg−1 support)

Wheat bran (66% 4 30 8.00±0.17 1.56±0.04 3.75±0.13 0.29±0.06 35 1260±30

water content) 4 20 4.33±0.01 0.56±0.03 4.53±0.11 0.21±0.01 52 530±30

4 10 0.28±0.01 0.30±0.02 1.90±0.03 0.23±0.01 38 390±20

Perlite (60% 6 30 9.15±0.24 0.14±0.01 1.13±0.06 0.09±0.01 10 nd4

water 6 20 4.47±0.27 0.12±0.01 1.11±0.08 0.08±0.01 13 nd

content) 3 10 6.50±0.19 0.09±0.01 0.53±0.03 0.09±0.01 31 nd

1The biotransformation yield (%) for L-phenylalanine is the ratio: (total moles of aryl metabolites (benzaldehyde, benzyl alcohol and benzoicacid) produced/ total moles of L-phenylalanine consumed)× 100.2Aryl alcohol oxidase.3Day of maximum production of benzaldehyde.4Not detected.

76

8

Table 3. Comparison between metabolites (benzyl alcohol, benzaldehyde and benzoic acid) production byB. adustain submerged liquid fermentation and solid statefermentation.

Submerged liquid fermentation (SLF)1 Solid state fermentation (SSF)(2)

Benzyl alcohol Benzaldehyde Benzoic acid Yield6 Benzyl alcohol Benzaldehyde Benzoic acid Yield

conc3 prod4 conc prod conc prod (%) conc prod conc prod conc prod %

(g kg−1) (g kg−1) (g kg−1) (g kg−1) (g kg−1) (g kg−1) (g kg−1) (g kg−1) (g kg−1) (g kg−1) (g kg−1) (g kg−1)

0.34±0.04(8)5 0.04 0.59±0(8)5 0.07 0.08±0(8)5 0.01 56 3.75±0.13(4)5 0.94 1.56±0.04(4)5 0.39 0.29±0.06(4)5 0.08 52

1Initial L-phenylalanine was 3 g kg−1 (Lapadatescuet al. 1997).2Initial L-phenylalanine was 30 g kg−1 support.3Conc= concentration of aryl metabolites.4Prod= productivity of aryl metabolites.5Numbers in parentheses show the day of maximum concentration of aryl metabolites.6The bioconversion yield is the ratio (total moles of metabolite formed/total moles of L-phenylalanine consumed)× 100.

769

References

Christen P, Villegas E, Revah S (1994) Growth and aroma produc-tion by Ceratocystis fimbriatain various fermentation media.Biotechnol. Lett.16: 1183–1188.

Christen P, Meza JC, Revah S (1997) Fruity aroma production insolid state fermentation byCeratocystis fimbriata: influence ofthe substrate type and the presence of precursors.Mycol. Res.101: 911–919.

De Jong E, Field JA, De Bont JAM (1994) Aryl alcohols in thephysiology of ligninolytic fungi. FEMS Microbiol. Rev.13:153–188.

Durand A, Renaud R, Maratray J, Almanza S, Diez M (1996)INRA-Dijon reactors for solid state fermentation: designs andapplications.J. Sci. Ind. Res.55: 317–332.

Gallois A, Gross B, Langlois D, Spinnler HE, Brunerie P (1990) In-fluence of culture conditions on production of flavor compoundsby 29 ligninolytic basidiomycetes.Mycol. Res.4: 494–504.

Gervais P, Sarrette M (1990) Influence of age of mycelium and wa-ter activity of the medium on aroma production byTrichodermaviride grown on solid substrate.J. Ferment. Bioeng.69: 46–50.

Glenn JK, Gold MH (1985) Purification and characterization of anextracellular Mn2+-dependenr peroxidase fromPhanerochaetechrysosporium. Arch. Biochem. Biophys.242: 329–341.

Gutiérrez A, Caramelo L, Prieto A, Martinez MJ, Martinez AT(1994) Anisaldehyde production and aryl-alcohol oxidase anddehydrogenase activities in ligninolytic fungi of the genusPleu-rotus. Appl. Environ. Microbiol.60: 1783–1788.

Hesseltine CW (1977) Solid-state fermentation: Part I.ProcessBiochem.12: 24–27.

Janssens L, De Pooter HL, Schamp NM, Vandamme EJ (1992) Pro-duction of flavours by microorganisms.Process Biochem.27:195–215.

Kawabe T, Morita H (1994) Production of benzaldehyde and benzylalcohol by the muxshroomPolyporus tuberasterK2606.J. Agric.Food Chem.42: 2556–2560.

Lapadatescu C, Feron G, Vergoignan C, Djian A, Durand A, Bonn-arme P (1997) Influence of cell immobilization on the productionof benzaldehyde and benzyl alcohol by the white-rot fungiBjerkandera adusta, Ischnoderma benzoinumand Dichomitussqualens. Appl. Microbiol. Biotechnol.47: 708–714.

Lapadatescu, C, Giniès, C, Djian, A, Spinnler, HE, Le Quéré,JL, Bonnarme, P (1999). Regulation of the synthesis of arylmetabolites by phospholipid sources in the white-rot fungusBjerkandera adusta.Arch. Microbiol. 171: 151–158.

Muheim A, Waldner R, Leisola MSA, Fiechter A (1990) An extra-cellular aryl-alcool oxidase from the white-rot fungusBjerkan-dera adusta. Enzyme Microb. Technol.12: 204–209.

Paszczynski A, Crawford R, Huynh VB(1988) Manganese perox-idase fromPhanerochaete chrysosporium:Purification. Meth.Enzymol.161: 264–268.

Roche N, Desgranges C, Durand A (1994) Study on the solid-state production of a thermostableα-L-arabinofuranosidase ofThermoascus aurantiacuson sugar beet pulp.J. Biotechnol.38:43–50.

Tien M, Kirk TK (1983) Lignin-degrading enzyme from the hy-menomycetePhanerochaete chrysosporiumBurds.Science221:661–663.