Introduction

Much attention has been given to xylan-degrading enzymes, including their potential industrial use in the animal feed, bread-making, and the paper and pulp industries (Bajpai 1999; Polizeli et al. 2005). More recently, particular attention has been given to the enzymes involved in enzymatic degradation of arabinoxylan in relation to “second generation” biofuel processes (Sørensen et al. 2005). Xylans consist of a linear backbone of β-1,4–linked d-xylopyranosyl units, which may carry various substitutions. The enzymatic degradation of xylans in turn requires action of different enzyme activities. β-Xylosidases (EC 3.2.1.37) attack the non-reducing ends of short xylooligosaccharides to liberate xylose and catalyse the cleavage of xylobiose (Rasmussen et al. 2006; Sørensen et al. 2003). β-Xylosidase activity is rate limiting in arabinoxylan hydrolysis (Poutanen and Puls 1988) and addition of pure β-xylosidase, purified from Trichoderma reesei, to a commercial, hemicellulolytic enzyme blend, “Ultraflo L” from Humicola insolens (Novozymes A/S, Bagsværd, Denmark), significantly boosts xylose liberation during enzymatic wheat arabinoxylan degradation (Sørensen et al. 2005).

Filamentous fungi are widely used as enzyme producers and are generally considered as more potent xylanase producers than bacteria and yeast (Haltrich et al. 1996; Polizeli et al. 2005). Strains of the genus Aspergillus have received particular attention due to their potential thermotolerance and production of thermostable enzymes (Castro et al. 1997). Thermostable enzymes are of interest because elevation of the reaction temperature (up to a certain limit) generally increases the reaction rate and reduces the risk of microbial contamination (Collins et al. 2005). Members of the Aspergillus section Nigri are particularly efficient producers of several types of extracellular enzymes (Samson et al. 2004; Serra et al. 2006). This group of fungi currently includes 17 species (Arbarca et al. 2004). This work was undertaken to identify novel xylan degrading enzyme activities in Aspergillus spp. with particular attention on identifying fungal strains producing thermally robust β-xylosidase activity.

Materials and methods

Chemicals

Wheat bran was from Cerealia Mills A/S [as per Sept. 2006 “Lantmännen Mills”] (Vejle, Denmark). Soluble wheat arabinoxylan was obtained from Megazyme (Bray, Ireland).

Microorganisms

The 20 aspergilli strains were all obtained from the IBT fungal culture collection at BioCentrum-DTU, Technical University of Denmark (Lyngby, Denmark). Table 1 shows the Aspergillus strains tested.

Table 1 Species from the genus Aspergillus evaluated in this work. Where known, the CBS number is also given
Full size table

Growth media

All fungi were initally cultivated on two different solid media to check growth and purity: Czapek-Dox yeast autolysate (CYA) agar and malt extract/agar (MEA) (Frisvad and Thrane, 2002). Afterwards, a basic liquid growth medium (GM) containing oat spelt xylan was used for the xylanase activity experiments, while a GM containing wheat bran was employed to screen for cellulolytic enzyme activities. The GM medium had the following basic composition: 3 g NaNO3 l−1, 0.5 g MgSO4 · 7H2O l−1, 1.3 g K2HPO4 · 3H2O l−1, 0.5 g KCl l−1, 10 mg FeSO4 · 7H2O l−1, 0.05 mg CuSO4 · 5H2O l−1 and 0.1 mg ZnSO4 · 5H2O l−1. For the cultivations for xylanase activity screenings, the oat spelt xylan (Sigma) was added to 30 gl−1. For the cellulolytic activity screenings the final wheat bran level in the GM was 600 gl−1. Prior to inoculation, the medium was autoclaved at 121°C for 15 min. After five days of incubation in the different growth media, the cultures were filtered through Whatman No. 4 filter paper (Whatman International Ltd., Maidstone, England).

Determination of xylanolytic and cellulolytic activities

Xylanolytic and cellulolytic activities were measured on filtered culture broths using Azo-wheat arabinoxylan and Azo-CM-Cellulose according to the procedures described by Megazyme (Bray, Ireland). Absorbances were measured at 590 nm, and the results were based on triplicate activity determinations on the filtered broth samples from each growth medium.

Determination of protein content

Protein levels were assessed directly on the filtered broths by the Lowry method using BSA as the standard.

Action on wheat arabinoxylan

Aliquots of soluble wheat arabinoxylan (1 mg) were dissolved in 1 ml deionised water. Each substrate sample was then incubated with 0.1 g enzyme protein·kg−1 dry wt wheat arabinoxylan in parallel evaluations for: (a) 2 h at 40°C; (b) 20 min at optimal temperature, i.e. 60°C for A. japonicus, and 75°C for A. niger and A. brasiliensis. After each incubation the samples were immediately heated at 100°C for 10 min to halt the enzyme reaction. The samples were then centrifuged at 20,000 g for ten minutes and the level of xylose liberated was determined by high performance anion exchange chromatography (Sørensen et al. 2003)

Determination of β-d-xylosidase activity

β-Xylosidase enzyme activity was measured by the p-nitrophenol method using 2.5 mM p-nitrophenol-β-d-xylopyranoside (Sigma) in 50 mM citrate buffer at pH 5 (Poutanen and Puls 1988). After incubation at 40°C for 20 min, the reaction was terminated by addition of 1 M Na2CO3 to a final concentration of 0.33 mM and the p-nitrophenol release from the substrate was measured at 400 nm. One katal of β-xylosidase activity per ml broth was calculated as equivalent to degradation of 1 mol β−1,4-xylosidic linkages s−1 as calculated from a standard curve of p-nitro-phenol. The results shown are averages of triplicate measurements made on each broth sample.

Statistics

The initial cellulolytic and xylanolytic activity data were analyzed by principal component analysis using The Unscrambler version 9.6 (Camo Software AS, Oslo, Norway). Significance of the β-xylosidase assay results was established at P ≤ 0.05. Differences in activities were determined by one-way analysis of variance with 95% confidence intervals calculated from pooled standard deviations (Minitab Statistical Software, Addison-Wesley, Reading, MA).

Results and discussion

Cellulolytic and xylanolytic activities of Aspergillus strains

Of the 20 strains tested in the screening work, A. brasiliensis (IBT−21946, CBS 101740, IMI 381727) isolated from Brazilian soil, A. japonicus (IBT−13519) isolated from soil in Burkina Faso, and A. niger (IBT-3250, CBS 554.65) isolated from tannin in Connecticut, USA, were found to have highest xylan-degrading activity and were selected for further study of their production of β-xylosidase activity.

pH optima

For all three of the selected A. niger, A. japonicus and A. brasiliensis strains, the highest β-xylosidase activities were at pH 5 (Fig. 1). For the A. niger strain, the β-xylosidase activities at pH 3 and 4 were equal to that at pH 5. For the A. niger and A. brasiliensis strains the lowest activities were at pH 6 whereas the β-xylosidase activity produced by A. japonicus at pH 6 was almost as high as that at pH 5. The A. japonicus activity was generally lower than that of the other two strains, however (Fig. 1).

Fig. 1
figure 1

pH optima of the β-xylosidase enzymes at dilution factor 2. β-xylosidase enzymes activity per ml broth was measured at pH 3 to 6 and at 40°C. A. niger (□), A. japonicus (○), A. brasiliensis (Δ)

Full size image

Temperature optimum

For the β-xylosidase activities produced by the A. niger and A. brasiliensis strains the optimal temperature was 75°C while the optimal temperature for the A. japonicus strain was at 60°C (Fig. 2). In general, these data for temperature and pH optima agree well with what has been reported previously by others (Table 2), but the temperature optima of the A. niger and A. brasiliensis strains studied here were clearly in the high end of the available data (Table 2). Previously, β-xylosidase activity produced by an A. niger strain was found to have temperature optimum above 75°C (Uchida et al. 1992), whereas, to the best of our knowledge, this is the first report of a high temperature optimum β-xylosidase activity production by A. brasiliensis and A. japonicus.

Fig. 2
figure 2

Temperature optima: β-xylosidase activity per ml broth was measured after incubation for 1 h at pH 5 at 40, 50, 60, 75 and 85°C. A. niger (□), A. japonicus (○), A. brasiliensis (Δ)

Full size image
Table 2 Properties of β-xylosidases produced by Aspergillus spp. Assay time is mentioned when known
Full size table

Thermal stability

Among the three strains, the thermal stability of the β-xylosidase activity produced by A. japonicus dropped rapidly above 40°C retaining 61% of the original activity after 1 h at 50°C, and only 16% at 60°C (Fig. 3). In contrast, 99% of the original β-xylosidase activity was retained after 1 h incubation of the A. niger culture broth at 60°C; after 1 h at 75°C the activity had dropped to 44% of the original activity (Fig. 3). The β-xylosidase activity produced by A. brasiliensis retained 62% of the activity after 1 h at 60°C, and 38% of the activity after 1 h at 75°C (Fig. 3). A β-xylosidase enzyme produced by T. reesei maintained only 20% of its activity at 70°C (Poutanen and Puls 1988), while a β-xylosidase produced by A. phoenicis maintained approximately 25% of the original enzyme activity after 1 h of incubation at 70°C while no activity was found after 1 h at 75°C (Rizzatti et al. 2001). Kiss and Kiss (2000) have reported that the β-xylosidase activity produced by A. carbonarius is completely lost after 30 min of incubation at 70°C, whereas Christov et al. (1999) found A. oryzae to produce β-xylosidase activity having a half life of 4 h at 60°C. The stability of the β-xylosidase activities produced by the A. niger IBT-3250 and A. brasiliensis IBT-21946 strains were thus better than the stability of previously studied β-xylosidases produced by strains of the genus Aspergillus or by Trichoderma reesei. A. brasiliensis is moreover a new specie in section Nigri which has the advantage of not being able to produce the mycotoxin ochratoxin A (Samson et al. 2004).

Fig. 3
figure 3

Temperature stability: β-xylosidase activity per ml broth was measured after incubation for 1 h at pH 5 and at each particular temperature

Full size image

Specific β-xylosidase activity

Among the three strains, A. brasiliensis had the highest specific β-xylosidase activity (Table 3). The underlying results indicated an almost similar protein production by A. niger and A. japonicus (0.56–0.62 mg protein/l), whereas the A. brasiliensis produced 85% and 70% more protein (1.05 mg/l) than A. niger and A. japonicus, respectively. When comparing the calculated specific β-xylosidase activities, A. brasiliensis moreover exhibited 66% higher specific activity than A. niger and 368% higher specific β-xylosidase activity than A. japonicus.

Table 3 Specific activity measured over 10 min at pH 5 and at optimal temperature; the optimal temperature given in parenthesis
Full size table

Action on wheat arabinoxylan

The data obtained on the genuine wheat arabinoxylan substrate (Table 4) verified the potency the A. brasiliensis β-xylosidase activity and confirmed the thermal activity of the A. niger and A. brasiliensis β-xylosidase enzymes on a polymeric xylan substrate.

Table 4 β-xylosidase activity a on polymeric wheat arabinoxylan
Full size table