the jowal of biouxical chemistry vol. 269, no. 6, issue of february 11, pp. 4299-4306 ... · 2001....

THE JOWAL OF BIOUXICAL CHEMISTRY 0 1994 by The American Society for Bioehemistry and Molecular Biology, Inc.

Vol. 269, No. 6, Issue of February 11, pp. 4299-4306, 1994 Printed in U.S.A.

2=O=Methyb~=mannose Residues are Immunodominant in Extracellular Polysaccharides of Mucor racemosus and Related Molds*

(Received for publication, September 16, 1993)

Gerhard A. De Ruiter$& Andrea W. Van Bruggen-Van der LugtS, Petra Mischnickn, Pieter Smidll, Jacques H. Van Boomll, Serve H. W. Notemans**, and Frank M. RomboutsS $$ From the $Department of Food Science, Wageningen Agricultural University, Bomenweg 2, 6703 HD Wageningen, The Netherlands, the Wepartment of Organic Chemistry, University of Hamburg, Martin-Luther-King Platz 6, 0-20146 Hamburg, Germany, the IIGorlaeus Laboratory, Department of Organic Chemistry, I? 0. Box 9502, 2300 RA Leiden, The Netherlands, and the **National Institute of Public Health and Environmental Protection (RIVM), Laboratory for Water and Food Microbiology, F! 0. Box 1, 3720 BA Bilthoven, The Netherlands

In this study, the structure of the immunodominant carbohydrate epitope of the extracellular polysaccharides from mold species belonging to the order Mucora- les reactive with rabbit IgG antibodies was elucidated. An exo-a-D-mannanase which was able to abolish the antigenicity of these polysaccharides completely was purified and characterized, and the activity was compared with that of an a-D-mannosidase. Analysis of the mono- meric reaction products after enzymatic treatment revealed the presence of 2-O-methyl-~-mannose residues. This compound is a constituent of the polysaccharides from the mold genera Mucor, Rhizopus, Rhizomucor, Ab- sidia, Syncephalastrum, and Thamnidium, and its occurrence in fungi has not been reported until now. Two mannan fractions which are highly reactive with rabbit IgG were isolated from the extracellular polysaccharides of Mucor racqmosus and characterized with ethylation analysis. The role of the newly found 2-O-methyl- D-mannose residues in the immunoreactivity was as- sessed by specific degradation of these mannans with the exo-a-n-mannanase and subsequent ethylation analysis. It was concluded that the immunodominant carbohydrates reactive with rabbit IgG are chains composed of a single terminal non-reducing 2-O-methyl-~-mannose residue, a(l-2)-linked to a short sequence of a(l-2)- linked D-mannose residues.

Molds belonging to the order of Mucorales (Zygomycetes), including the genera Mucor, Rhizopus, Rhizomucor, Syncepha- lustrum, Absidia, and Thamnidium have a worldwide distribu- tion. Some species are pathogenic for crop plants and many can be involved in food spoilage resulting in huge economic loss (1). In medicine, some species are important as the cause of mu- cormycosis (zygomycosis; 2, 3). An increasing interest is being shown for the antigenic properties of extracellular polysaccharides (EPSs),l which are excreted by all tested species of this

*This work was supported by the Netherlands’ Foundation for Chemical Research (SON) with financial aid from the Netherlands’ Technology Foundation (STW) and by the Deutsche Forschungsgemein- schafi. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked ”advertisement” i n accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Corporate Research, P. 0. Box 414, 3770 AK Barneveld, The Nether- 5 Present address: Hercules European Research Center (HERC), Div.

lands. $$To whom correspondence and reprint requests should be ad-

dressed. Fax: +31-8370-84893. The abbreviations used are: EPSs, extracellular polysaccharides;

order (4, 5), as target compounds for modern immunological detection of these species in food, feed, and clinical medical samples (3, 6, 7).

It was revealed that immunoglobulin G (IgG) antibodies obtained after immunization of many rabbits with either polysaccharides isolated from mycelium or with EPSs of different species of the order Mucorales were monospecific and reacted with all species belonging to this order (5, 8-10). This indicates the presence of a common immunodominant epitope reactive with all of these rabbit-raised antibodies, which structure is cur- rently not known in detail. If a mouse was used for immunization with these EPSs, a monoclonal antibody has been obtained with similar specificity, but apparently reactive with a different carbohydrate epitope. This difference was explained by consid- ering that both a different host animal and a different immunization procedure were used (11).

The polysaccharides of several species of Mucorales have been characterized partly, and it was shown that the rabbit IgG antibodies reacted with the mannose- and fucose-containing fractions (9, 12-20). Polymers of p(l-4)-linked D-glucuronic acid residues have also been isolated from the EPSs, but these were not immunoreactive (16, 18, 21). Two highly immunoreactive fractions from Absidia cylindrospora, a fucomannopeptide and a mannoprotein, were isolated by Yamada et al. (19, 20). The mannose residues were mainly (1-2)- and (14-linked as shown by methylation analysis (18-20). These authors assumed an important role for cY(14-linked mannose oligomers as the common antigenic determinant of Mucorales (9, 17,20), while other reports point to a(l-2)-mannose linkages (18). However, these antibodies did not show any reactivity with mannans obtained from yeasts such as Candida albicans and Saccharomyces cerevisiae, which are mainly composed of a( 1- 2)- and cY(l-61-linked mannose residues (5, 9, 10).

To explain these apparent contrasting results, the structure of the residues which are immunodominant in mucoralean polysaccharides has to be studied with more refined methods. In all of the above mentioned studies, the methylation analysis procedure was used to characterize the glycosidic linkages of the antigenic carbohydrate fractions of these molds. A disad- vantage of this commonly used method is that methyl-ethers of sugars escape detection, as it is not possible to distinguish between the naturally present -0-methyl groups and -0-methyl

IgG, immunoglobulin G; HPAEC, high performance anion-exchange

GLC, gas-liquid chromatography; DAS-ELISA, double-antibody sand- chromatography; HPLC, high performance liquid chromatography;

wich-enzyme-linked immunosorbent assay; FPLC, fast protein liquid chromatography.

4299

4300 2-0-Me-mannose Is Immunodominant in Mucoralean Molds

groups which are obtained by chemical substitution during methylation analysis. To avoid this problem, ethylation analysis is a suitable alternative. Furthermore, purified enzymes can be used as sophisticated tools for the elucidation of immunodominant carbohydrates, as recently shown for p( l-5)-linked D-galactofuranoside oligomers from the EPSs of Penicillium and Aspergillus molds (22).

In the present paper, the structure of the immunodominant epitope of mucoralean EPSs reactive with rabbit-raised antibodies is established. An exo-a-D-mannanase, which was able to abolish the antigenicity of these EPSs completely, was purified and characterized. Furthermore, two highly immunoreactive mannans were isolated from the EPS of Mucor racernosus and characterized using ethylation analysis. The role of the newly found 2-O-methyl-~-mannose residues in the immunoreactivity of these EPSs was established by synthesis of two oligosaccha- rides and subsequent treatment with different enzymes. Fi- nally, the structure of these immunodominant carbohydrates was revealed.

MATERIALS AND METHODS Isolation and Purification of EPSs-Extracellular polysaccharides of

the strains Mucor hiemalis Wehmer CBS 201.28, M. racemosus Fres. H473-R5, M. circinelloides van Tieghem M 40, Rhizopus stolonifer (Ehrenb.) Lind CBS 609.82, Rhizomucorpusillus (Lindt) Schipper CBS 432.78, Absidia corymbifera (Cohn) Sacc. & Trotter LU 017, Syncepha- lastrum racemosum Cohn CBS 443.59, Thamnidium elegans Link CBS 342.55 were produced, isolated, and purified by ethanol precipitation as described (12). Mannans from the yeasts C . lambica (Lindner and Ge- noud) Van Uden and Buckley PB l B, Hansenula holstii Wickerham CBS 2028, Pichia membranaefaciens (Hansen) Hansen CBS 107, and Saccharomyces exiguus Reess ex Hansen WM 5 were isolated as described (10).

Antibodies-Four batches of antibodies were obtained by immunization of three rabbits with EPS from M. racemosus (numbers 14, 1000, and 1201) and one rabbit with EPS from M. circinelloides (number 15). Immunization of the rabbits and purification of the IgG fraction from the sera was performed as described by Notermans and Heuvelman (7).

Enzyme Preparations-The epitope-degrading exo-a-D-mannanase was purified from a crude enzyme preparation of fungal origin (E-icho- derma harzianum), commercially available as Glucanex (Novo Nordisk Ferment AG, Dittingen, Switzerland). An a-D-mannosidase (EC 3.2.1.24) isolated from Jack beans (Canavalia ensiformis) was obtained from Sigma (M 7257) and used without further purification.

Periodate Oxidation-Periodate oxidation was performed on a polysaccharide sample of 5 mg by treatment with 2 ml of a solution of 50 nm NaI04 in 0.25 M formic acid, adjusted to pH 3.7, for different incubation times at 4 "C as described recently (11).

Determination of the Sugar Composition and Protein Content-The carbohydrate composition of EPS fractions was determined after liberation of the respective sugar residues by methanolysis combined with trifluoroacetic acid hydrolysis as described recently (23). Subsequently, the amount of liberated monosaccharides was estimated with the use of high performance anion-exchange chromatography (HPAEC) using a Dionex Bio-LC HPLC system (Dionex, Sunnyvale, CA) equipped with a CarboPac PA 100 column (4 x 250 mm) and pulsed-amperometric detection (23).

Protein was determined by the colorimetric method of Sedmak and Grossberg (24), with bovine serum albumin as the standard.

Ethylation Analysis-Approximately 1 mg of the sample was dissolved in 150 pl of dimethyl sulfoxide in a V-vial. Freshly prepared 1.5 M lithium dimethylsulfinyl anion (62 pl) was added through the septum cap, and the mixture was stirred for 2 h at room temperature. To the ice-cooled solution, 7.5 p1 of ethyliodide (Merck, Darmstadt, Germany) was added, and the mixture was stirred for 2 h at room temperature. Ethylation was repeated twice, but the third time 22.5 pl of ethyliodide was used. The perethylated polymeric products were isolated by dialy- sis and subsequent extraction with dichloromethane.

One-third of the sample was hydrolyzed with 2 M trifluoroacetic acid at 120 "C for 2 h. After evaporation of the acid, the residue was reduced with 0.5 M sodium borodeuteride in 2 M ammonia for 1 h at 60 "C. The excess of sodium borodeuteride was destroyed with acetic acid, and the borate was subsequently removed by repeated evaporation with acidic methanol. The residue was acetylated with acetic anhydride and pyri-

dine at 90 "C for 2 h. The reaction mixture was washed with a solution of saturated NaHC03, and the products were extracted with dichloromethane. Identification and quantification of the partially ethylated alditol acetates was performed by one- and two-dimensional GLC-mass spectrometry as described previously (22).

Another part of the sample was subjected to reductive cleavage which was performed according to the method of Bowie et al. (25). To a solution of 0.2 mg of the perethylated mannan in 50 pl of dichloromethane, 1615 equivalenb'glycosidic bond of triethylsilane and trimethylsilyl trifluoromethanesulfonate were added. After incubation for 20 h at room temperature, 10 pl of acetic anhydride were added. After 2 h the reaction was quenched with a saturated solution of NaHC03, and the products were extracted with dichloromethane. GLC-mass spectrometry was performed with a VG/76250S instrument as described (22), and ammonia was used as reactant gas for chemical ionization.

DAS-ELISA-A double-antibody sandwich (DAS)-ELISA was carried out by first binding the antibodies to the wall of a microtiter plate, followed by the addition of the polysaccharide sample to the wells, and finally by using the same antibodies conjugated to peroxidase. The DAS-ELISA was camed out using polyvinyl microtiter plates using 3,3',5,5'-tetramethyl benzidine as the peroxidase substrate (12). The ELISA reactivity was expressed as the minimal detectable concentration of the antigenic material which is defined as the concentration in distilled water just giving a positive reaction, 2.e. an extinction at 450 nm >0.1 above that of a blank, containing no antigenic material.

Purification of the Epitope-degrading Exo-a-D-mannanase-The purification was carried out at 4 "C, and all buffers contained 0.01% (w/v) sodium azide to prevent microbial growth. Five g of Glucanex were dissolved in 10 ml of 10 m~ sodium acetate (pH 5.0), and the solution was centrifuged to remove solids. The epitope-degrading enzyme was isolated and purified from this preparation using anion-exchange chromatography and adsorption chromatography with an hydroxyapatite column according to the procedure as described previously for the purification of an exo-PD-galactofuranosidase (22). The final purification step (Fig. 1) was performed on a fast protein liquid chromatography (FPLC) system (Pharmacia LKB Biotechnology, Uppsala, Sweden) equipped with a Mono S HR 515 cation-exchange column (50 x 5 mm). Elution was done with sodium succinate buffer (20 nm (pH 4.0)) and a sodium chloride gradient from 0 to 0.1 M, at a flow rate of 1 ml/min. During gradient elution, peak control was used to elute protein peaks with a minimum amount of contamination by maintaining the composition of the eluent at a fixed value during elution of the peaks. The epitope-degrading activity of different enzyme fractions was measured by incubation of EPS from M. racemosus and the respective enzyme fractions followed by a DAS-ELISA. Incubation was performed in a reaction mixture containing 175 pl of EPS (1 mg/ml) in 50 nm sodium acetate (pH 5.5) and 25 pl of an enzyme fraction. After incubation for 16 h at 30"C, the enzyme was inactivated by heat treatment (5 min, 100 "C) prior to the DAS-ELISA. The purity of the fractions containing the epitope-degrading enzyme was checked by SDS-gel electrophoresis.

SDS-gel Electrophoresis, Isoelectric Focusing, and Titration Curve -Determination of the molecular mass was performed by SDS-gel electrophoresis on a 1615% polyacrylamide gradient gel using the Phast- System (Pharmacia LKB Biotechnology) according to the instructions of the supplier. Standards in the range of 10 to 100 kDa were used for calibration. The determination of the isoelectric point, isoelectric focusing, and titration curves were performed on homogeneous polyacrylamide gels containing Pharmalyte carrier ampholytes, which generate a linear pH gradient from 3 to 9 in the gel. Standards with an isoelectric point in the range of 3 to 9 were used. Proteins were detected by Coomassie Brilliant Blue R-250 and silver staining.

Determination of the Enzyme Activity, lkmperature, and pH Opti- mum-The activity of the purified exo-a-o-mannanase fraction (12.5 pl) and the a-D-mannosidase from Jack beans (12.5 pl) was expressed in milliunits calculated after incubation with 175 pl(1 mg/ml) of EPS from M. racemosus at 30 "C in 12.5 pl of 2 M sodium acetate buffer (pH 5.0). One milliunit was defined as the amount of enzyme able to release 1 nmol of mannose from this EPS preparation/min at 30 "C and pH 5.0. The digests were subsequently analyzed for the amount of mannose released after incubation for 1 and 2 h, using HPAEC as described above. Optimum temperature and optimum pH were determined with 350 pl of a solution of EPS (1 mg/ml) from M. racemosus buffered with 25 pl of 2 M sodium acetate with a pH varying from 3 to 6. These solutions were incubated with 0.45 milliunit (25 pl containing 125 ng of protein) purified exo-a-D-mannanase for 2 h at different temperatures between 20 and 60 "C. After inactivation of the enzyme (5 min, 100 "c) the released monosaccharides were measured by HPAEC.

Substrate Specificity of the Enzymes-Enzyme activity toward vari-

2-0-Me-mannose Is Immunodominant in Mucoralean Molds 430 1 ousp-nitrophenyl-glycosides (Sigma) was measured spectrophotometri- cally at 405 nm, using the molar extinction coefficient 13.700 M-l Cm". In addition, differently linked D-mannose oligomers were used as a substrate. The dimers Manal+2Manal-,Me (dimer l), Manal-, 6Manal-Me (dimer 2), the trimers 2-O-Me-Mana1-,2Manal- 2MancrljMe (trimer 1) and 2-O-Me-Mana1-,6(2-O-Me)Manal-6(2- 0-Me)Manal+Me (trimer 2), and the tetramer Manal-,BManal- 2Manal+2Manal+Me were synthesized using the iodonium ion-me- diated glycosylation procedure of properly protected ethyl 1-thio-a-D- mannopyranosides in which N-iodosuccinimide and trifluoromethane sulfonic acid (catalytic) are used as promotor, as described recently (26). The non-methylated 3-linked D-mannose dimer was purchased (Sigma M 8897). An aliquot of 50 pl of the respective oligomers (2 mg/ml) was diluted with 135 pl of 150 l ~ l ~ sodium acetate (pH 5.0) and incubated with 0.75 milliunit of the respective enzyme for 16 h at 30 "C. The amount of o-mannose and 2-O-methyl-~-mannose released was determined by HPAEC as described.

The mannan-degrading activity of the enzymes was also tested by incubation with four different yeast mannans. Incubation was performed using 175 pg of purified mannan in 250 pl of 80 m sodium acetate buffer (pH 5.0) and 0.75 milliunit of the respective enzymes at 30 "C for 16 h. After inactivation of the enzymes, the products were analyzed as described above. Glucanase side activities were measured after incubation of the enzymes with different glucans as described (22).

Anion-exchange Chromatography of the EPSs-This was performed on a column (12 x 1.6 cm) of DEAE-Sepharose CL-GB (Pharmacia), equilibrated with 0.05 M sodium acetate buffer (pH 5.0). After loading of the sample of EPS (4 ml of a 2.5 mg/ml solution), the column was washed with 25 ml of buffer and then eluted at 30 mVh with a linear gradient (100 ml) of 0.05-1 M sodium acetate buffer, followed by 50 ml of 1 M buffer. Fractions (3 ml) were assayed for neutral sugars and glucuronic acid by the automated orcinol method and automated m- hydroxydiphenyl method, respectively (12).

Size-exclusion Chromatography ofthe EPSs-Size-exclusion chromatography was performed on a column (91 x 1.5 cm) of Bio-Gel P-10 (200-400 mesh; Bio-Rad) eluted with 0.1 M sodium acetate buffer (pH 4.0) at 8 ml/h. Samples of polysaccharides were dissolved in 2 ml of this buffer and loaded onto the column. Fractions of 1.9 ml were collected and analyzed on their neutral sugar content using the automated orcinol method as described above. In order to pool the collected fractions optimally, the molecular weight distributions and corresponding DAS- ELISA activities were checked using size-exclusion chromatography in the high performance mode as described (12).

Enzymatic Degradation ofthe EPSs-An aliquot of 175 pl of polysaccharide solution (1 mg/ml) and 10 pl of 2 M sodium acetate buffer (pH 5.0) was incubated with 0.75 milliunit (65 p l ) of the exo-a-D-mannanase or a-mannosidase at 30 "C for 16 h. The immunoreactivity of the native and enzyme-treated polysaccharides was measured using the DAS- ELISA. f i r inactivation (5 min, 100 "C) the remaining polysaccharides were isolated by removing the reaction products by using a Microsep filter (Filtron Technology Corp., Northborough, MA) with a cutoff of 10 kDa. The reaction products were analyzed using HPAEC as described above and by GLC-mass spectrometry after derivatization to alditol acetates. The latter were prepared by transferring the filtrate to a screw-cap tube which was evaporated to dryness by a stream of air (25 "C). The monosaccharides released by the enzyme were first reduced to alditols with NaBH, and subsequently converted into their corresponding alditol acetates using 1-methyl imidazole and acetic anhydride (23, 27). Finally, ethylacetate was added and the solution was transferred to a vial and sealed. The alditol acetates were analyzed by GLC-mass spectrometry using a Hewlett-Packard 5890-5970 MSD equipped with a Chrompack CP Si1 19 CB (Middelburg, The Nether- lands) capillary column (26 m, inner diameter 0.22 mm, film thickness 0.18 pm) with a temperature gradient of 160-250 "C, 2 Wmin.

RESULTS Antibodies-Four batches of antibodies obtained after immu-

nization of three rabbits with EPSs from M. racemosus and one with EPSs of M. circinelloides were tested thoroughly on their reactivity with EPS preparations of 39 different yeasts (10) and 24 different fungi (10, 28). The four batches of antibodies reacted similarly in all cases and were specific for the EPSs of the mold species of the order Mucorales tested, with the exception of the species belonging to the genus Mortierella sensu stricto (e.g. M. polycephala, M. reticulata, M. hyalina; 28). As the antibodies all cross-reacted with the yeast Pichia membranae-

0

n a

07

I A A ""

4' A I'

110 120 130 140 1SO- elution volume lml 1

FIG. 1. Final purification step of the epitope-degrading exo-a- mmannanase on a mono S cation-exchange column using a FPLC. A sodium succinate buffer was used for elution with a sodium chloride gradient. Symbols: -,A,,,; - - - - -, NaCl concentration of the effluent. Fractions I and I1 were pooled as indicated.

faciens only (lo), these batches of rabbit IgG antibodies were considered to be identical. As a result thereof, it is most likely that a common characteristic structure of the mucoralean EPSs is recognized by these antibodies, similar as previously shown for the mold species belonging to the genera Penicillium and Aspergillus (22). Therefore, this structure is highly immunodominant and is called "the epitope" in the rest of this paper, although it is likely that more epitopes are present on these EPSs, however, with a much lower immunogenicity in rabbits.

Periodate Deatment of the EPSs-As the EPS preparations of Mucorales contain both carbohydrate and protein segments, the antigenicities were studied by periodate treatment. Sugar residues of glycoproteins can specifically be oxidized by mild periodate treatment of carbon atoms with vicinal hydroxyls leaving the protein part unaffected (29). The EPS preparations were treated with 50 ~ l l ~ sodium periodate for periods of 15 min, 1 h, 8 h, 24 h, and 6 days. The immunoreactivity of the EPSs before and after periodate treatment was established by DAS-ELISA. In all cases, a complete loss of immunoreactivity of the EPSs was observed after periodate treatment (results not shown). Sugar analysis before and after periodate treatment revealed that the galactose, mannose, and glucose residues were destroyed by periodate treatment, whereas the fucose residues were hardly affected (results not shown).

Purification of the Epitope-degrading Exo-a-D-mannanase "Forty crude enzyme preparations were checked on their abil- ity to degrade the antigenic determinants of EPSs from Muco- rales, and an enzyme preparation isolated from the fungus Dichoderma harzianum was found to contain this activity. The enzyme did not bind to the DEAE Bio-Gel A anion-exchange column and was not retained on a hydroxyapatite column as described (22). The non-adsorbing fraction of the latter column was separated with a Mono S/HR cation-exchange column using the FPLC as shown in Fig. 1. To point out the protein fractions which were able to degrade the epitopes of M. race- mows EPS, each fraction eluting from this column was incubated with this EPS and the immunoreactivity was measured with DAS-ELISA. The fractions which possessed epitope-degrading activity (Z and ZZ, Fig. 1) were checked by isoelectric focusing, and two enzyme fractions could be isolated. The isoelectric points of the two fractions were 6.6 and 7.0, respectively, and their molecular masses measured by SDS-gel electrophoresis were identical (50 ma). The titration curves, indicating the punty and the electrophoretic mobility of the protein at different pH values, revealed homogeneous enzymes as illustrated for fraction I in Fig. 2.

Activity of the Enzymes toward Various Substrates-The en-

4302 2-0-Me-mannose Is Immunodominant in Mucoralean Molds

line o f - application

I I I I I

8 7 6 5 4 PH

FIG. 2. Titration curve of the purified epitope-degrading ex- a-wmannanase (I) using a pH gradient from 3 to 9. The protein on the gel was stained with silver.

zyme fractions I and I1 obtained from cation-exchange chromatography (Fig. 1) showed identical activities toward the various substrates tested. As these fractions only differed in isoelectric point, they were considered to be isoenzymes and called epitope-degrading exo-a-D-mannanase in the rest of this paper. The enzyme did not show any activity toward thep-nitrophenyl derivatives of the pyranose residues of a- and P-D-mannose, p-D-glucose, a-L-fucose, and P-D-glucuronic acid, even after incubation for 24 h. The epitope-degrading exo-a-D-mannanase was active toward dimer 1 (Manal+2Manal+Me), and the corresponding tetramer showed a slight activity toward the 3-linked dimer of D-mannose but was not active on the dimer 2 (Manal+GManal+Me) of D-mannose. Opposed to this, the a-D-mannosidase from Jack beans showed activity toward the p-nitrophenyl derivative of a-D-mannose and was also able to hydrolyze the glycosidic linkages of all the a-linked D-mannose dimers and the tetramer tested.

Both the a-D-mannosidase and our exo-a-D-mannanase released similar amounts of mannose after incubation with the yeast mannans tested. The epitope-degrading exo-a-D-mannanase contained traces of glucanase activity as the enzyme was able to release minor amounts of glucose from glucans, while the a-D-mannosidase did not contain this activity. How- ever, no release of glucose was observed aRer incubation with the mucoralean EPS preparations, therefore, no attempts were made to remove this glucanase side activity.

Enzymatic Degradation of the EPSs of Mucoralean Molds- The action of the purified exo-a-D-mannanase on the EPSs of Mucorales was compared with the action of the commercially available a-D-mannosidase. The immunoreactivity of eight EPS preparations originating from species belonging to six genera within the order Mucorales was determined both before and after treatment with the enzymes by DM-ELISA. As shown in Table I, treatment of the EPSs with the a-D-mannosidase did not affect their immunoreactivity. Opposed to this, incubation with the exo-a-D-mannanase resulted in a complete disappear- ance of the immunoreactivity of the mucoralean EPSs as in most cases the minimal detectable quantity of EPSs, before enzyme treatment (<lo to 1000 ng/ml) increased to more than 10 pg/ml after treatment (Table I).

The reaction products obtained after degradation of the mucoralean EPSs were analyzed with HPAEC as illustrated for M. circinelloides in Fig. 3. Mannose was released with a-D-mannosidase (Fig. 3 a ) and with the epitope-degrading exo-a-D-mannanase (Fig. 3b) as is apparent from the peak with a retention time of 12.5 min in this chromatographic system. The amount

Immunoreactivity of mucoralean EPSs tested with a DAS-ELISA TASLE I

before and after treatment (16 h, 30 "C, pH 5.0) with a-D-mannosidase or the epitope-degrading exo-a-D-mnnnalurse

Experiments were done in triplicate.

Origin of EPS Native ARer treatment with

a-o-Mannosidase Exo-a-o-mannanase

M. racemosus M. hiemalis M. circinelloides R. stolonifer R. pusillus A. corymbifera S. mcemosum T elegans

++a

++ ++++ +++ +++ ++ ++ ++

+ ++

++++ +++ +++ ++ ++ ++

tion in DAS-ELISA ++++, c10 ng/ml; +++, 10-100 ng/ml; ++, 100-1000 a Immunoreactivity expressed as the minimal detectable concentra-

ng/ml; +, 1-10 pg/ml; -, >lo pg/ml.

b.

0 10 20 30 i o S O 60

the reaction products obtained after incubation of EPS from M. FIG. 3. High performance anion-exchange chromatogram of

circinelloides with the a-wmannosidase (a) and with the epitope-degrading ex-a-wmannanase (b ) . The main peak with a retention time of 12.5 min represents mannose. PAD, pulsed-amperometric detection.

retention time Iminl

of mannose liberated from the mucoralean EPSs by the exo-a- D-mannanase was in all cases somewhat higher than the amount released by a-D-mannosidase, except in the case of M. racemosus, as shown in Table 11. The EPS preparations were also incubated first with the a-D-mannosidase and, after removal of the liberated mannose residues, with the exo-a-D- mannanase, which resulted in the liberation of additional mannose in all cases (Table 11).

The HPAEC chromatogram of the reaction products of the epitope-degrading enzyme (Fig. 3b) revealed the presence of a characteristic second peak with a retention time of approximately 8.3 min. This compound was released after enzymatic treatment of all EPS preparations of Mucorales tested with the epitope-degrading exo-a-D-mannanase enzyme but not with the a-D-mannosidase. I t was not possible to identify the structure of this compound with the use of HPAEC technology. Compounds which are expected to elute in this region using these HPAEC conditions are arabinose, methyl-ethers of sugars, and the 2-de- oxy-2-aminosugars (23, 30).

GLC-Mass Spectrometry Analysis of the Enzyme Reaction Products-To characterize the unknown compound with a retention time of 8.3 min on the HPAEC system (Fig. 3b), the EPS reaction products obtained with the epitope-degrading exo-a-Do- mannanase were isolated by filtration. This filtration step resulted in a slight loss of material probably by binding to the filter material. After separation of the products released by the enzymes, these were converted into the corresponding alditol acetates after reduction of the aldehyde function on the C-1 to an alcohol, without any further chemical hydrolysis step. The reaction mixture of the EPS preparation of M. hiemalis was

2-0-Me-mannose Is Immunodominant in Mucoralean Molds 4303

analyzed by GLC-mass spectrometry, resulting in the chromatogram shown in Fig. 4, with mannose (retention time 21.7 min) as the main component. The mass spectrum of the unknown compound (inset) with a retention time of 19.3 min was identified as belonging to the alditol acetate of a 2-0-methyl-hexose residue (31). As mannose, galactose, and glucose were the only aldose sugars present (231, the mass spectrum shown in Fig. 4 can originate from the alditol acetates of 2-0-methyl ethers of these sugar residues, as their alditol acetates only differ in their retention time on the GLC column (31). Therefore, the retention time of the characteristic unknown compound was compared with the retention times of the alditol acetates of the 2-0-methyl ethers of mannose, galactose, and glucose using GLC columns with different polarity (results not shown). This unequivocally revealed that the presence of both 2-0-methyl- glucose and 2-0-methyl-galactose could be excluded. In conclusion, the epitope-degrading exo-a-D-mannanase released both mannose and 2-0-methyl-mannose residues from M. circinelloides and M. hiernalis EPSs. The amount of 2-0-methyl-mannose residues which was liberated by this enzyme was determined by HPAEC for all mucoralean EPS preparations available as listed in Table 11. Quantification was performed using the molar response factor of mannose (23). No attempts

TABLE 11 Products liberated upon degradation of EPSs isolated from molds belonging to the order of Mucorales with a-D-mannosidase, the epitope-

degrading em-a-o-mannanase, and the two enzymes consecutively

Origin of EPS

M. racemosus M. hiemalis M. circinelloides R. stolonifir

A. corymbifera R. pusillus

S. racemosum T elegans

a-D-Mannosidase, Man"

8.1 (63) 2.1 (14) 2.4 (7) 2.8 (29) 5.7 (37) 0.6 (8)

11.5 (55) 1.6 (19)

Exo-a-D-mannanase

2-0-Me- Manb

7.0 (54) 0.1 3.2 (20) 0.2 6.7 (19) 0.9 3.2 (33) 0.3 6.2 (41) 0.3

15.1 (73) 0.3 1.1 (14) 0.4

2.4 (28) 0.2

mannanase &er E x o - a - ~

a-o-mannosidase, additional Man'

0.2 0.4 0.6 0.2 0.4

<o. 1 0.6 0.3

a Expressed in pg released from 175 pg of EPS used in the incubation

pressed as relative to the initial amount of mannose present in these experiments. The values in parentheses represent the amount ex-

EPSs (from Ref. 23). Values are the average of duplicate experiments. Amount of 2-0-methyl-mannose residues liberated from 175 pg of

EPS expressed in pg, calculated with the relative response factor of mannose (23).

e Expressed in pg, EPS (175 pg) was first incubated with a-o-mannosidase and aRer removal of the reaction products by filtration sub-

rected for slight losses due to filtration. sequently incubated with exo-a-o-mannanase. Values were not cor-

chromatogram of the reaction prod- FIG. 4. GLC-mass spectrometry

ucta obtained after incubation of EPS from M. hiemah with the epitope-degrading exo-a-wmannanase after derivatization to alditol acetates. The mass spectrum of the compound with a retention time 19.3 min, which is as- signed to 2-O-methyl-ma~ose, is given in the inset. The main peak with a retention time of 21.7 min represents mannose.

100

80 0

c1

0 al

al

F 60

a

$- 40 0 d

E 20

0

were made to reveal the absolute configuration (D or L) of the 2-0-methyl-mannose residues.

Zsolation of the Immunoreactive Fractions from EPS of M. racernosus-EPS of M. racernosus was separated by DEAE anion-exchange chromatography in three main fractions similar as shown for the EPS of M. hiernalis by De Ruiter et al. (12). The immunoreactivity of the fractions was determined by DAS- ELISA, which revealed that only fractions I and I1 were reactive with the rabbit IgG antibodies raised against EPS of M. racernosus.

The sugar composition of the fractions from the DEAE anion- exchange column is listed in Table 111. Both fractions I and I1 were mainly composed of mannose residues. Fraction I11 from this column was not reactive with the rabbit IgG and may be identical to mucoran described by Bartnicki-Garcia and Lind- berg (32).

The immunoreactive fractions I and I1 from the DEAE anion- exchange column were separated further by size-exclusion chromatography using a P-10 column, as shown in Fig. 5. The most immunoreactive fractions were pooled as indicated in Fig. 5 (horizontal bars), dialyzed, and lyophilized and are referred to as mannans I and 11, respectively.

Characterization of Mannans Z and ZZ-The immunoreactivity of both mannans from Mucor racernosus was determined using DAS-ELISA. The minimal detectable concentration of mannan I was approximately 10 ng/ml whereas the minimal concentration of mannan I1 which could be detected in this ELISA was 10 times higher. As shown in Table 111, the immunoreactive mannans I and I1 were mainly composed of mannose residues, with low amounts of glucose, fucose, and galactose and in case of mannan I1 also some glucuronic acid. Further-

T-LE I11

fluoroacetic acid hydrolysis of the fractions of EPS from M. racemosus Sugar composition determined after methanolysis and subsequent tri-

obtained after separation by DEAE anion-exchange chromatography and further with P-10 size-exclusion chromatography

Values are the average of duplicate experiments.

Carbohydrate composition"

Fuc 2-O-Me-Manb Man Gal Glc GlcA Fraction

DEAE I 6 1 84 1 7 DEAE I1

1 10 1 79 2 4

DEAE I11 27 4

9 11 2 51 Mannan I 5 2 82 2 9 0 Mannan I1 7 1 80 2 8 2

a Sugar composition expressed in mole percentages. ' 2-0-Methyl-mannose residues. The amount was calculated by tak-

ing the relative molar response factor of mannose (23).

"\ 4 6 8 10 12 ld 16 18 20 22 24

re tent ion t ime (min)

4304 2-0-Me-mannose Is Immunodominant in Mucoralean Molds 500

- 400 1.0

- 200- .

- z 100-

g o -

2 .- L

300-

200-

100-

n -

al u

L " I O 20 30 i 0 SO 60

fractions (1.9mlI

FIG. 5. Separation of the fractions I (a) and I1 (b ) obtained from DEAE anion-exchange chromatography of the EPS from M. rmem8u8 (12) by size-exclusion chromatography (P-10 column). The line represents the colorimetric sugar determination performed on each fraction and the bars the ELISA extinction of the different fractions.

more, approximately 142% of 2-0-methyl-mannose residues was present in both mannans. The carbohydrate content of the mannans I and I1 was 88 and 84%, respectively. Only traces of protein could be detected in these mannans (results not shown).

Both mannans were examined with ethylation analysis showing that both were mainly constituted of (l-a)-linked mannose residues (33-35%), terminal mannose residues (23- 25%), (1-6)-linked mannose residues (12-16%), differently branched mannose residues, and terminal 2-0-methyl-mannose residues (1-2%). The latter compound was apparent from the mass spectra of the degradation products of the perethylated mannans. The molecular masses of the products obtained by standard ethylation analysis and by reductive cleavage were determined by chemical ionization mass spectrometry and were 365 and 262, respectively. From their electron impact mass spectra (Fig. 6), the nature of the 2-0-methyl-mannose component was unequivocally proved by identification of 1,5- di-O-acetyl-3,4,6-trii-O-ethyl-2-O-methyl-mannitol (Fig. 6 a , standard ethylation analysis) and 1,5-anhydro-3,4,6-tri-O- ethyl-2-0-methyl-mannitol (Fig. 6b, reductive cleavage).

The most characteristic primary fragments of a-cleavage of the identified open-chain 2-0-methyl-mannitol (Fig. 6a) were those resulting from fragments with m / z 247 (CH20ethyl- HCOAc-HCOethyl-HCOethyl), m / z 189 (CH20ethyl-HCOAc- HCOethyl, C4-Cs), and m / z 176 (CHDOAc-HCOmethyl-HCO- ethyl). The most characteristic fragments of the anhydro-2-0- methyl-mannitol obtained after reductive cleavage (Fig. 6b) were m / z 262 (M'), m / z 216 (M' minus ethylOH from position 3), m / z 203 (M' minus CH20ethyl), and m / z 186 ( m l z 216 minus CH,O), as revealed by comparison with the mass spectra of 1,5-anhydro-2,3,4,6-tetra-O-methyl-mannitol (33) and 2-0- acetyl-l,5-anhydro-3,4,6-tri-O-methyl-mannitol (34). Since the compounds of Fig. 6 were the only 2-0-methyl-mannose derivative present in both of these mannans, it was concluded that 2-O-methyl-~-mannose occurs only as non-reducing terminal residues.

Enzymatic Degradation of the Mannans-To elucidate those parts of the mannans which are the epitopes reactive with the rabbit IgG antibodies, mannans I and I1 were treated with the two enzymes, a-D-mannosidase and exo-a-D-mannanase, as described above. As shown in Table IV, both enzymes released similar amounts of mannose residues, whereas the immunoreactivity of the mannans was hardly affected after treatment with a-D-mannosidase, and abolished completely after incubation with exo-a-D-mannanase. The latter enzyme also released 2-O-methyl-~-mannose residues.

Y i' relative abundance

0 2$1 a. 50 100 150 200 250 300 350

1007 115

59 99 1-

11s

0 b.

50 100 150 200 250 300 350 m/z

FIG. 6. Electron impact-mass spectra of l,S-di-O-acetyl-3,4,& tri-0-ethyl-2-0-methyl-mannitol (a) and l,S-anhydro-3,4,&tri-O- ethyl-2-0-methyl-mannitol (b). The first spectrum was obtained after standard ethylation analysis and the latter after reductive cleavage

the EPSs of all mold species belonging to the order Mucorales. of the non-reducing terminal 2-0-methyl-mannose residues present in

TABLE IV Products liberated upon &gradation of the highly immunoreactive mannans I and II isolated from EPS from M. racemosus and their

remaining DAS-ELISA reactivity See legend Table I.

Mannan Native a-D-Mannosidase Exo-a-o-mannanase

Man" ELISA Man 2-0-Me Manb ELISA

I ++++ 18 +++ 17 0.3 I1 ++ 14 ++ 13 0.2

- -

a Amount of mannose expressed in pg released from 175 pg of the

2-0-Methyl mannose residues. The amount was calculated by tak- respective mannans after enzyme incubation.

ing the relative molar response factor of mannose (23).

The remaining mannans were isolated by filtration after treatment with either of the two enzymes, and their composition was determined by ethylation analysis. This revealed that the amount of a(l-21-linked mannose residues decreased con- siderably in both cases. Furthermore, the relative amounts of 2-O-methyl-~-mannose residues increased after a-D-mannosidase treatment whereas this compound could not be detected anymore after treatment with the exo-a-D-mannanase (results not shown).

Enzymatic Degradation of 2-O-Methyl-~-mannose-containing Oligomers-The above mentioned results suggested that the carbohydrate residues involved in the epitopes of mucoralean polysaccharides has to be interlinked in such a way that vicinal hydroxyl groups remain and that the non-reducing terminal position is occupied by a 2-O-methyl-~-mannose residue. The two possible structures which fulfill these requirements, the trimers 1 and 2, were synthesized by Smid et al. (26) and contain both a non-reducing terminal 2-O-methyl-~-mannose residue. Both were incubated with the exo-a-D-mannanase, and the monomers released were detected with HPAEC. After incubation for 16 h, trimer 1 (2-O-Me-Mana1-2Manal-, 2Mana1-Me) was almost completely degraded by the exo-a+- mannanase resulting in monomers of 2-O-methyl-~-mannose and D-mannose residues with a relative proportion of 1:2, respectively, and trimer 2 (2-O-Me-Manal+6(2-O-Me)Manal- 6(2-O-Me)Manal+Me) was not degraded at all (results not shown).

DISCUSSION In this study, the structure of the immunodominant carbo-

hydrates of EPSs of M. racemosus and related molds which are uniquely responsible for the antigenic properties of these EPSs in rabbits was revealed. Mild periodate oxidation resulted in the complete loss of ELISA reactivity with the rabbit IgG antibodies and in the loss of some sugar residues. As amino acids of glycoproteins can only be oxidized by periodate if higher

2-0-Me-mannose Is Immunodominant in Mucoralean Molds 4305

concentrations and longer incubation times are used (291, it can be assumed that the immunoreactivity of these EPS preparations is based on the integrity of sugar residues with vicinal hydroxyl groups.

An enzyme which was able to abolish the immunoreactivity of these polysaccharides was purified to homogeneity from a crude preparation of T. harzianum. This enzyme acted as an exo-enzyme as only monomers were released from yeast mannans and from fungal EPSs. However, no activity was found on the p-nitrophenyl derivative of a-D-mannose, indicating that the enzyme needs more than 1 sugar residue for binding. Al- though the absolute configuration of the released products was not determined, it is very likely that the enzyme only released monomers with the D-configuration as it was able to degrade a( l-a)-linked oligomers synthesized from D-mannose. There- fore, this enzyme should be called an exo-a-D-mannanase (35) as opposed to the a-D-mannosidase from Jack beans which did react with the p-nitrophenyl-a-o-mannoside.

Two immunoreactive mannans were isolated from the crude EPS preparation of M. racemosw, which only represent a few percent of the initial EPS. Mannan I was immunochemically most reactive, contained 40% (1-21-linked mannose residues, and is probably similar to the mannoprotein fraction isolated previously from EPS ofA. cylindrospora (18). Mannan I1 was 10 times less immunoreactive than mannan I and was composed of mannose, fucose, and a small amount of glucuronic acid. This fraction is probably similar to the fucomannopeptide as described by Yamada et al. (18). Both mannans contain l-2% of 2-O-methyl-~-mannose residues, apparently present as non- reducing terminals only. Specific degradation of these mannans with the exo-a-D-mannanase and analysis of the remaining mannans by ethylation analysis revealed that the remaining not-immunoreactive mannan was devoid of 2-O-methyl-~-mannose residues indicating that these residues are immunodominant. As shown in this study, the a-o-mannosidase cleaves all the

freely exposed differently linked a-D-mannose residues but did not influence the immunoreactivity of EPSs of Mucorales. The exo-a-D-mannanase liberated both terminal a-O-methyl-~-mannose residues and terminal mannose residues from ~ ~ 4 1 - 2 ) - linked chains of mannose with or without a terminal 2-0- methyl+-mannose residue. Incubation of the EPS preparations with a-D-mannosidase followed by exo-a-D-mannanase showed that the second enzyme liberated additional mannose residues. Therefore, it was concluded that the mannose residues which are liberated aRer subsequent exo-a-D-mannanase treatment originate from a( 1-21-linked D-mannose chains which arise after removal of the 2-O-methyl-~-mannose residues. As the syn- thetic trimer 1 was degraded completely by the exo-a-D-mannanase, and trimer 2 could not be degraded at all, it was confirmed that the epitope of mucoralean EPSs carries a 2-0- methyl-D-mannose residue at the non-reducing terminal, and is further composed of a( l-a)-linked D-mannose residues, as pro- posed in Fig. 7.

0-Methyl ethers of sugars are characteristic constituents of lipopolysaccharides from photosynthetic bacteria (36, 37). The 2-0-methyl-mannose residue has been determined in the cya- nobacterium Synechocystis (38) and in the nitrogen-furing bacteria of the genus Frankia (39). In addition, 2-0-methyl-mannose has been reported in soil a h r hydrolysis of raw humus (40). In fungi, the 0-methyl ethers of sugars are less common, but the 3-0-methyl ether from mannose has been reported as an important compound of the conidial walls of Coccidioides immitis (36, 41). To our knowledge, the occurrence of 2-0- methyl-D-mannose in fungi has never been reported before.

The presence of the terminal 2-O-methyl-~-mannose residue at the mucoralean epitope explains why no cross-reactions of

FIG. 7. Model for the immunodominant carbohydrates of extracellular polysaccharides of molds belonging to the order Mu- corales reactive with rabbit-raised IgG antibodies. The epitope is composed of a single-terminal non-reducing 2-O-methyl-~-mannose residue a(l-2)-linked to a short sequence of a(l-2)-linked D-mannose residues.

IgG antibodies raised against EPSs from Mucorales were observed with mannans from the yeast species of the genera Saccharomyces and Candida (5,9, 10). As the characterization of the epitope in the studies of Miyazaki et al. (9, 17) and Yamada et al. (18-20) was mainly based on methylation analysis, which do not allow for the detection of %O-rnethyl-~-mannose residues, it is most likely that 2-O-methyl-~-mannose containing oligomers were characterized erroneously as containing mannose residues only.

In conclusion, the purified exo-a+-mannanase was a very useful tool for the elucidation of the immunodominant carbohydrate epitope of mucoralean EPSs reactive with rabbit IgG. The use of the enzyme combined with analysis of the mono- meric reaction products by both HPAEC and GLC-MS and the remaining polysaccharide by ethylation analysis allowed us to prove the structure of the immunodominant carbohydrates of these EPSs as being a 2-O-methyl-~-mannose terminally linked to the 2-position of a sequence of a-~-(l-2)-mannose residues (Fig. 7). With this knowledge, these antibodies can be used to develop rapid and reliable immunoassays for detection of mold species belonging to the genera Mucor, Rhizopus, Rhizomucor, Absidia, Syncephalastrum, and Thamnidium in foods, agricultural products, and medical samples (6).

Acknowledgment-We gratefully acknowledge Dr. H. J. Kamphuis (Gerkens Cacao Industrie b.v., Wormer, The Netherlands) for critical reading of the manuscript.

REFERENCES 1. Pitt, J. I., and Hocking, A. D. (1985) Fungi and Food Spoilage, pp. 143-167,

2. Sugar, A. M. (1992) Clin. Infect. Dis. 14, Suppl. 1, S12€-S129 3. Kaufman, L., Turner, L. F., and McLaughlin, D. W. (1989) J. Clin. Microbiol.

27, 1979-1982 4. De Ruiter, G. A,, Smid, P., Van der Lugt, A. W., Van Boom, J. H., Notermans,

S . H. W., and Rombouts. F. M. (1991) in Fungal Cell Wall and Immune Response (Latg6, J.-P., and Boucias, D., eds) pp. 169-180, NATO AS1 Series H 53, Springer Verlag, Berlin

5. Hayashi, O., Yadomae, T., Yamada, H., and Miyazaki, T. (1978) J. Gen. Micro- biol. 108,345-347

6. De Ruiter, G. A., Notermans, S . H. W., and Rombouts, F. M. (1993) Dends

7. Notermans, S. , and Heuvelman, C. J. (1985)Int. J. Food. Microbiol. 2,247-258 Food. Sci. lkchnol. 4,91-97

8. Hearn, V. M., and Mackenzie, D. W. R. (1980) Mol. Immunol. 17, 1097-1103 9. Miyazaki, T., Yadomae, T., Yamada, H., Hayashi, O., Suzuki, I., and Ohshima,

Y. (1980) in Fungal Polysaccharides (Sandford P. A., and Matsuda K., eds) pp. 81-94, American Chemical Society, Symposium Series 126, Washington D. C.

10. De Ruiter, G. A., Van Bruggen-van der Lugt, A. W., Nout, M. J. R., Middel-

(1992) Antonie van Leeuwenhoek 62, 189-199 hoven, W. J., Soentoro, P. S. S . , Notermans, S . H. W., and Rombouts. F. M.

11. De Ruiter, G. A., Van Bruggen-van der Lugt, A. W., Bos, W., Notermans, S . H. W., Rombouts, F. M., and Hofstra, H. (1993) J. Gen. Microbiol. 139, 1557- 1564

12. De Ruiter, G. A., Van der Lugt, A. W., Voragen, A. G. J., Rombouts, F. M., and Notermans, S . H. W. (1991) Carbohydr. Res. 216.47-57

13. Hayashi. 0.. Yamada, H., Yadomae, T., and Miyazaki, T. (1978) J. Gen. Micro- biol. 108, 289-295

14. Miyazaki, T., and Irino, T. (1971) Chem. Pharm. Bull. (Tokyo) 19, 145&1454 15. Miyazaki, T., and Irino, T. (1972) Chem. Pharm. Bull. (Tokyo) 20,33&335

Academic Press, New York

4306 2-0-Me-mannose Is Immunodominant in Mucoralean Molds 16. Miyazaki, T., Suzuki, I., Yamada, H., Yadomae, T., Kumazawa, Y., and Mizu-

noe, K. (1977) Kitasato Arch. Exp. Med. SO, 1-11 17. Miyazaki, T., Hayashi, O., Ohshima, Y., and Yadomae, T. (1979) J. Gen. Mi-

crobiol. 111,417422 18. Yamada, H., Ohshima, Y., and Miyazaki, T. (1982) J. Gen. Microbwl. 128,

189-197 19. Yamada. H., Ohshima, Y., and Miyazaki, T. (1982) Chem. Pharm. Bull. (Imkyo)

SO, 1784-1791 20. Yamada, H., Ohshima, Y., and Miyazaki, T. (1982) Carbohydr: Res. 110, 113-

126 21. De Ruiter, G. A., Josso, S . L., Colquhoun, I. J., Voragen, A. G. J., and Rombouts,

E M. (1992) Carbohydc Polym. 18,l-7 22. Van Bruggen-van der Lugt, A. W., Kamphuis, H. J., De Ruiter, G. A., Misch-

6096-6102 nick, P., Van Boom, J. H., and Rombouts, E M. (1992) J. Bacteriol. 174,

23. De Ruiter, G. A,, Schols, H. A,, Voragen, A. G. J., and Rombouts, F. M. (1992)

24. Sedmak, J. J., and Grossberg, S . E. (1977)Anal. Biochem. 79,544-552 Anal. Biochem. 207, 176-185

25. Bowie, J. U., Trescony, P. V., and Gray, G. R. (1984) Carbohydr: Res. 125, 301-307

26. Smid, P. (1993) Use of Monosaccharides in the Preparation of Biologically Active Compounds, Ph.D thesis, pp. 35-56, State University Leiden, The Netherlands

27. Englyst, H. N., and Cummings, J. H. (1984) Analyst 109.937-942 28. De Ruiter, G. A,, Van Bruggen-van der Lugt, A. W., Rombouts, F. M., and Gams,

29. Clamp, J. R., and Hough, L. (1965) Biochem. J. 94, 17-24 30. Lee, Y. C. (1990) Anal. Biochem. 189,151-162 31. Jansson, P.-E., Kenne, L., Liedgren, H., Lindberg, B., and Liinngren, J. (1976)

A Practical Guide to the Methylation Analysis of Carbohydrates, Vol. 8, University of Stockholm, Stockholm

W. (1993) Mycol. Res. 97,690-696

32. Bartnicki-Garcia, S., and Lindberg, B. (1972) Carbohydc Res. 23,7=5 33. Kemper, M. S . (1988) Ph.D thesis, University of Minnesota, Minneapolis, MN 34. Van Langenhove, A,, and Reinhold, V. N. (1985) Carbohydr Res. 143, 1-20 35. Snaith, S . M., and Levvy, G . A. (1973) Adv. Carbohydr: Chem. Biochem. 28,

36. Lindberg, B. (1990) Adv. Carbohydr: Chem. Biochem. 48,279-318 37. Weckesser, J., Drews, G., and Mayer, H. (1979) Annu. Rev. Mierobiol. 33. 38. Sehmidt, W., Drews, G., Weckesser, J. and Mayer, H. (1980) Arch. Microbiol.

39. Mort,A., Normand, P., and Lalonde, M. (1983) Can. J. Microbiol. 29,993-1002 40. Ogner, G. (1980) Soil Sci. 129, 1-4 41. Cole, G. T., and Kirkland, T. N. (1991) in The Fungal Spore and Disease

Initiation in Plants and Animals (Cole, G. T., and Hoch, H. C., eds) pp. 403-443, Plenum Press, New York

401445

215239

127, 217-222

the jowal of biouxical chemistry vol. 269, no. 6, issue of february 11, pp. 4299-4306 ... · 2001....

Documents