|Home | About | Journals | Submit | Contact Us | Français|
Paenibacillus sp. W-61 is capable of utilizing water-insoluble xylan for carbon and energy sources and has three xylanase genes, xyn1, xyn3, and xyn5. Xyn1, Xyn3, and Xyn5 are extracellular enzymes of the glycoside hydrolase (GH) families 11, 30, and 10, respectively. Xyn5 contains several domains including those of carbohydrate-binding modules (CBMs) similar to a surface-layer homologous (SLH) protein. This study focused on the role of Xyn5, localized on the cell surface, in water-insoluble xylan utilization. Electron microscopy using immunogold staining revealed Xyn5 clusters over the entire cell surface. Xyn5 was bound to cell wall fractions through its SLH domain. A Δxyn5 mutant grew poorly and produced minimal amounts of Xyn1 and Xyn3 on water-insoluble xylan. A Xyn5 mutant lacking the SLH domain (Xyn5ΔSLH) grew poorly, secreting Xyn5ΔSLH into the medium and producing minimal Xyn1 and Xyn3 on water-insoluble xylan. A mutant with an intact xyn5 produced Xyn5 on the cell surface, grew normally, and actively synthesized Xyn1 and Xyn3 on water-insoluble xylan. Quantitative reverse transcription-PCR showed that xylobiose, generated from water-insoluble xylan decomposition by Xyn5, is the most active inducer for xyn1 and xyn3. Luciferase assays using a Xyn5-luciferase fusion protein suggested that xylotriose is the best inducer for xyn5. The cell surface Xyn5 appears to play two essential roles in water-insoluble xylan utilization: (i) generation of the xylo-oligosaccharide inducers of all the xyn genes from water-insoluble xylan and (ii) attachment of the cells to the substrate so that the generated inducers can be immediately taken up by cells to activate expression of the xyn system.
Xylan is a major component of hemicellulose in plant cell walls, and it is the second most abundant polysaccharide, next to cellulose, on earth (44). The widespread occurrence of xylanolytic enzymes hydrolyzing β-1,4-xylan backbones and side chain sugars of hemicellulose among microorganisms (18) reflects the importance of this abundantly available carbohydrate as a carbon and energy source. Endo-1,4-β-xylanases (xylanase; EC 220.127.116.11) that cleave the β-1,4-glycoside bonds of the xylan backbones are classified into seven families of glycoside hydrolases (GH) based on their amino acid sequences (4). Xylanases of different GH families can synergistically hydrolyze xylan in vitro (4), and many microorganisms that produce xylanases from different GH families appear to cooperate with each other in the degradation of xylan (6, 41). Some microorganisms, such as Aspergillus spp., Bacillus spp., Clostridium spp., Paenibacillus spp., and Streptomyces spp., that can utilize xylan produce several xylanases of different GH families (18, 27, 29).
Paenibacillus sp. W-61 was isolated from soil as a Gram-positive bacteria that can degrade and utilize xylan as the sole source of carbon and energy via a multixylanase system (42). Five xylanases (Xyn1 to Xyn5) have been identified in W-61 culture supernatants growing on xylan, and these have been characterized (8, 16, 26, 30, 42). Xyn1 and Xyn3 are secreted into the medium, where the latter enzyme is proteolytically converted into Xyn2 (26). Recently, we found a cytoplasmic membrane protein, LppX, which serves as a chaperone during secretion of Xyn1 across the cytoplasmic membrane (12). Xyn5 is detected in the cell envelope during the exponential growth phase, released into the medium during the stationary phase of growth, and subsequently cleaved by proteolysis into Xyn4 (16). Xyn5 hydrolyzed the water-insoluble fraction from xylan 3.5-fold and 5.4-fold more efficiently than Xyn3 (8, 30) and Xyn1 (unpublished data), respectively.
Xyn5 has a molecular mass of 140 kDa and consists of a catalytic domain of GH family 10, domains similar to carbohydrate-binding modules (CBMs) of the superfamily 22 CBM (CBM22), a family 9 CBM (CBM9), including a lysine-rich domain homologous to scaffoldin-dockerin-binding protein A (SdbA) of the Clostridium thermocellum cellulosome, and S-layer-homologous (SLH) motifs (16). CBM9 and CBM22 of Xyn5 have been shown to display cellulose binding abilities (16). Some members of the low-GC content, Gram-positive bacteria have cell wall-anchored xylanases with modular structures similar to the structure of Xyn5 (21). Clostridial xylanases are cell wall bound within large enzyme complexes called cellulosomes, possibly allowing the bacterial cell to utilize substrate more efficiently as greater functional display is possible (5, 7, 11). Cell surface-localized xylanases of these organisms may be viewed as candidates for the efficient conversion of water-insoluble xylan to xylo-oligosaccharide products. However, the role of cell surface-localized xylanase in the multixylanase expression system for water-insoluble xylan utilization has not been investigated.
In this study, we demonstrated that Xyn5 molecules covered the entire cell surface and were associated with the cell wall fraction through a C-terminal SLH domain. Cell surface-localized Xyn5 appears to play essential roles in water-insoluble xylan utilization by activating expression of a multienzyme system for efficient hydrolysis of water-insoluble xylan and releasing oligosaccharides from water-insoluble xylan that act as inducers for all xyn genes, including xyn5. The function of Xyn5 domains in conjunction with a proposed role as inducers or generators of a xylan degradation system in Paenibacillus sp. W-61 is discussed.
Table Table11 lists the bacterial strains and plasmids used in the present study. Paenibacillus sp. W-61 and its derivatives were cultured in water-insoluble xylan medium, containing 0.7% (wt/vol) water-insoluble xylan as a carbon source (or other specified carbon source) in minimal medium (medium I) (42), for growth tests, total RNA preparation, and cellular protein preparations for Western blotting and zymography. Water-insoluble xylan and water-soluble xylan were prepared as follows (24, 43). Ten grams of oat spelt xylan (Nakalai Tesque, Japan) was suspended in 100 ml of distilled water, and the suspension was stirred overnight at 4°C. After centrifugation (at 20,000 × g for 20 min at 4°C), the resultant supernatant was filtered through a GA-200 (Advantec, Japan) glass filter paper and lyophilized to obtain the water-soluble xylan fraction. The pellet was also lyophilized and used as the water-insoluble xylan fraction. Luria-Bertani (LB) medium (17) was also used for preparation of chromosomal DNA and plasmid DNA transformations. Escherichia coli strains were routinely cultured in LB medium. Where appropriate, antibiotics were added to cultures at the concentrations indicated below. Paenibacillus and E. coli were cultured at 30°C and 37°C under aerobic conditions, respectively, unless otherwise noted.
Chromosomal DNA of Paenibacillus sp. W-61 and plasmid DNA were prepared as previously described (30). The PCR was carried out using conditions reported previously (32). A Δxyn5 mutant (PW5001) was constructed as follows: an internal 2.5-kb DNA fragment (nucleotides [nt] 19 to 2361 relative to the translation initiation codon) of xyn5 was amplified by PCR, using Paenibacillus sp. W-61 chromosomal DNA as a template, with primers P1 and P2 (Table (Table2)2) (32). After the 3′ overhangs created during PCR had been removed with T4 DNA polymerase (Blunting Kination Ligation kit; TaKaRa Bio, Japan), the amplified xyn5 fragments were cloned into the SmaI site of pUC119 (TaKaRa Bio, Japan). An internal 902-bp HindIII region of xyn5 was replaced with a chloramphenicol acetyltransferase (cat) fragment, which had been PCR amplified from plasmid pC194 (14) using primers P3 and P4 (Table (Table2).2). The protruding ends of both the plasmid and the 839-bp DNA fragments containing cat were removed as above. The resultant xyn5::cat fragment was amplified by PCR using primers P1 and P2, the 3′ overhangs were removed, and the fragment was inserted into the SmaI site of the shuttle vector pKAF, which contains the temperature-sensitive ori of pE194, for replication in Gram-positive bacteria (35). Plasmid pX5001 (Table (Table1)1) was introduced into Paenibacillus sp. W-61 by electroporation (2), and the resultant transformant was then cultivated in LB broth containing chloramphenicol (Cm; 10 μg/ml) at 43°C. After appropriate dilution, the culture was spread on LB plates containing 10 μg/ml Cm to select for the Δxyn5 mutant strain PW5001 (xyn5::cat) that had acquired the mutant gene via a single crossover upstream of xyn5, as verified by Southern blotting (26).
A plasmid carrying xyn5, used in complementation tests, was constructed by amplifying the xyn5 coding region (nt −107 to 4143, relative to the translation initiation codon) by PCR using W-61 chromosomal DNA as a template and primers P5 and P6 (Table (Table2).2). The amplified DNA fragment was digested with SmaI and BamHI and cloned into plasmid pHY300PLK between the corresponding sites to yield plasmid pX5002 (Table (Table1).1). A region of xyn5 (nt −107 to 3450) was amplified by PCR using plasmid pUX5S-22 (Table (Table1)1) as a template and primers P5 and P7 (Table (Table2)2) to construct a plasmid containing xyn5 Δslh. The 3′ overhangs were removed from the amplified DNA fragments and inserted into the SmaI site of plasmid pHY300PLK to obtain plasmid pX5003 (Table (Table11).
The luciferase (luc) coding sequence (nt 1 to 1909) of the PicaGene Basic Vector 2 (TOYO B-Net, Japan) was amplified by PCR using primers P8 and P9 (Table (Table2).2). After the protruding ends were converted to blunt ones, the amplified DNA fragment was cloned between the HindIII and EcoRI sites of pHY300PLK, resulting in plasmid pX5004 (Table (Table1).1). A putative promoter region of xyn5 (nt −216 to 36) including a Shine-Dalgarno sequence was amplified by PCR from W-61 chromosomal DNA using primers P10 and P11 (Table (Table2)2) and cloned as a KpnI-SmaI fragment into plasmid pX5004 (Table (Table1)1) between the corresponding sites, to obtain plasmid pX5005 (Pxyn5-luc) (Table (Table11).
An SLH domain-encoding region (nt 2631 to 4075) was amplified by PCR using W-61 chromosomal DNA as a template and primers P12 and P13 (Table (Table2).2). The amplified DNA fragments were cloned as SmaI-BamHI fragments into plasmid pGEX4T-1 between the corresponding sites to yield plasmid pX5006 (Table (Table11).
Xyn5 was labeled with immunogold by the method of Egelseer et al. (9, 10). Early-stationary-phase cells of strain W-61 and the Δxyn5 mutant (PW5001) were grown in 1 ml of medium I containing 0.7% (wt/vol) water-soluble xylan and harvested by centrifugation. The cell pellets were washed once with phosphate-buffered saline (PBS). Following resuspension in 250 μl of PBS, cells were incubated with 2 μl of anti-Xyn5 antiserum diluted 10-fold with PBS at 4°C for 10 h. The cells were washed once and resuspended in 10 μl of protein A-colloidal gold conjugates (GE Healthcare, United Kingdom), 10 nm in diameter, and incubated at room temperature for 1 h. Free protein A-colloidal gold conjugates were removed by washing the cells three times with 250 μl of PBS. The immunogold-stained cells were resuspended in 10 μl of PBS. Approximately 2 μl of the cell suspensions was applied onto glow-discharged carbon-coated copper grids and observed under a Hitachi Z-8100 electron microscope (Hitachi, Japan) operating at 75 kV.
E. coli BL21(DE3) cells harboring pX5006 were incubated and grown at 30°C in LB medium containing 50 μg/ml ampicillin. When the optical density at 600 nm (OD600) reached 0.5, isopropyl-β-d-thiogalactopyranoside was added to a final concentration of 1 mM. After incubation for an additional 4 h, the cells were harvested by centrifugation at 4°C and resuspended in 10 ml of 50 mM sodium phosphate buffer (buffer A; pH 7.3) and then disrupted by passage through a French pressure cell at 12,000 lb/in2. After unbroken cells and large debris were removed by centrifugation at 3,500 × g for 10 min at 4°C, the supernatant was centrifuged at 200,000 × g for 60 min at 4°C to remove cell envelopes. The resultant supernatant was loaded onto a GSTrap FF column (GE Healthcare, United Kingdom), and glutathione S-transferase (GST)-SLH fusion protein was eluted using 10 mM reduced glutathione.
The peptidoglycan (PG) of E. coli DH5α and the Δxyn5 mutant was prepared as follows. Cell wall components were prepared from E. coli DH5α and Δxyn5 mutant (PW5001) cells using the method described by Ries et al. (28). A 100-ml culture of Δxyn5 mutant was incubated at 30°C in LB medium until the optical density at 600 nm reached 1.0. The cells were harvested by centrifugation at 4°C, resuspended in 20 ml of buffer A, and then twice disrupted by passage through a French pressure cell at 12,000 lb/in2. After unbroken cells and large debris were removed by centrifugation at 3,500 × g for 10 min at 4°C, the supernatant was centrifuged at 200,000 × g for 60 min at 4°C to obtain cell envelopes. The resultant cell envelope (4 g, wet weight) was boiled for 2 h in 50 ml of buffer A containing 4% (wt/vol) sodium dodecyl sulfate. After centrifugation at 100,000 × g for 60 min at 20°C, the pellets were washed three times with buffer A, and then 400 mg of pellet was suspended in 20 ml of buffer A and treated with trypsin (0.1 mg/ml) from bovine pancreas (Wako Pure Chemicals, Japan) at 37°C for 2 h to obtain the cell wall fraction. To obtain the PG, the trypsin-treated cell wall fraction was resuspended in 5 ml of buffer A containing 5% (wt/vol) trichloroacetic acid (TCA) and incubated at 35°C for 24 h. After centrifugation at 20,000 × g at room temperature for 10 min, the pellets were washed three times with buffer A and used as a PG fraction. To prepare the secondary cell wall polymer (SCWP) of the Δxyn5 mutant, the trypsin-treated cell wall fraction of the Δxyn5 mutant was digested with 500 μl of 3 mg/ml lysozyme (Seikagaku Co., Japan) in buffer A at 37°C for 1 h. The pellets were collected by centrifugation at 20,000 × g at 4°C for 10 min, washed three times with buffer A, and used as an SCWP fraction. Purity of PG and SCWP was checked using 12.5% (wt/vol) SDS-PAGE gels, and sugar concentrations were quantified using the method of Hodge and Hofreiter (13) with d-glucose as a standard.
Purified PG or SCWP (2 mg each) was mixed with 3 to 5 μg of GST-SLH in 400 μl of 50 mM sodium phosphate buffer (pH 7.0). The mixtures were incubated at 4°C for 60 min with occasional stirring, and then the precipitates were collected by centrifugation at 20,000 × g at 4°C for 10 min and resuspended in 400 μl of the same buffer. From this suspension 10 μl was analyzed by SDS-PAGE and stained with Coomassie brilliant blue. Signal intensities of protein bands were quantified using Image J software (http://rsbweb.nih.gov/nih-image/).
Extracellular proteins from 2 ml of culture were precipitated with 10% (wt/vol) TCA, washed with cold acetone, and dried with a centrifugal vacuum evaporator. The dried protein samples and cells were dissolved in 200 μl of SDS-PAGE sample buffer (19), and 10-μl aliquots were resolved by SDS-PAGE. Proteins on polyacrylamide gels were electrically transferred onto Hybond enhanced chemiluminescence (ECL) membranes (GE Healthcare, United Kingdom). Xyn1, Xyn3, and Xyn5 on the membranes were detected with antibodies against the relevant proteins (16, 26) and with an ECL detection kit (GE Healthcare, United Kingdom) according to the supplier's protocols.
For zymography, both extracellular proteins and cells in the 2-ml culture aliquot were concentrated by TCA precipitation and resolved by SDS-PAGE. After the SDS-PAGE gel was washed three times in 2.5% (vol/vol) Triton X-100, it was rinsed once in 50 mM sodium phosphate buffer (pH 7.0). The renatured gel was plated onto a 1.2% (wt/vol) agarose gel containing 0.5% (wt/vol) Remazol brilliant blue (RBB)-xylan (Sigma-Aldrich, Japan) as a substrate and incubated at 37°C for 6 h (30).
Total xylanase activity in the culture was measured using 2% (wt/vol) water-soluble xylan as a substrate under conditions described previously (43). One unit of enzyme activity was defined as the amount of enzyme that liberated 1 μmol of reducing sugar, as xylose, per minute from the substrate (43). The reducing sugars were measured by the methods of Somogyi (36).
To assay luciferase activity, cells in the 2-ml culture aliquot were centrifuged and washed once with 50 mM sodium phosphate buffer (pH 7.0). The cells were incubated in 100 μl of the same buffer containing 30 μg of lysozyme at 37°C for 10 min. The cell lysate was centrifuged for 10 min at 5,000 × g and 4°C to obtain soluble cellular proteins. The luciferase assay was carried out using a PicaGene assay system (TOYO B-Net, Japan). The reaction mixture, containing 20 mM Tricine, 1.07 mM Mg(CO3)4 Mg(OH)2 5H2O, 2.67 mM MgSO4, 0.1 mM dithiothreitol (DTT), 270 μM coenzyme A, 470 μM luciferin, 530 μM ATP, and the supernatant of the cell lysate, was incubated at 20°C for 30 min. The amount of emitted luminescence was measured using a Luminescencer PSN AB-2200 (ATTO, Japan). Promoter activity was measured as relative light units (RLU) and all data were normalized against the 50 mM sodium phosphate buffer (pH 7.0).
Water-insoluble xylan (10 mg) was incubated with purified Xyn5 (2 U) in 2 ml of 50 mM sodium phosphate buffer (pH 7.0) at 37°C for 2 h. The remaining water-insoluble xylan was removed by centrifugation at 20,000 × g for 30 min at 4°C. The supernatant was twice filtered through number 5C filter paper (Advantec, Japan) and lyophilized. Xylo-oligosaccharides from water-insoluble xylan were separated by high-performance liquid chromatography (HPLC) as described previously (40, 43, 45). The hydrolyzed products and purified xylo-oligosaccharides were analyzed by thin-layer chromatography using silica gel 60F 254 (Merck) (16).
Total RNA was prepared from W-61 and Δxyn5 mutant (PW5001) cells by the hot-phenol method described by Aiba et al. (1). cDNAs of xyn1, xyn3, and xyn5 were synthesized using avian myeloblastosis virus reverse transcriptase XL (TaKaRa RNA PCR kit, version 3.6; TaKaRa Bio, Japan) in 10-μl reaction mixtures containing 100 ng of total RNA and 1.0 μM concentrations of primers P14, P15, and P16 (Table (Table2).2). Quantitative PCR was conducted using a LightCycler (Roche, Germany). Reaction mixtures (20 μl) contained 5 μl of cDNA and 0.1 μM each of primers P17 and P18 for xyn5, P19 and P20 for xyn1, or P21 and P22 for xyn3 (Table (Table2)2) and 2 μl of Faststart DNA polymerase (LightCycler Faststart DNA Master SYBR green I kit; Roche, Germany). The PCR cycling conditions consisted of an initial denaturation step at 95°C for 10 min, followed by 45 cycles of denaturation at 95°C for 10 s, annealing at 55°C for 5 s, and extension at 72°C for 10s. The experiment was done in triplicates, and all data were normalized to the amount of xyn transcripts at 0 h.
Nucleotide sequencing of the ΔxynS mutant and the plasmid constructed in this study (Table (Table2)2) was performed on an ABI Prism 310 genetic analyzer (Applied Biosystems).
Previously, we found that Xyn5 located in the envelope fraction from Paenibacillus sp. W-61 cells grown in water-insoluble xylan medium (16). The susceptibility of Xyn5 to trypsin and the existence of fragments liberated in the supernatants by the trypsin treatment indicate that Xyn5 is exposed on the cell surface (16). To investigate the role of cell-bound Xyn5 on water-insoluble xylan utilization, we constructed the Δxyn5 mutant, PW5001 (xyn5::cat). Immunostaining for Xyn5 with anti-Xyn5 antibodies and immunogold particles under a transmission electron microscope (TEM) revealed clusters of immunogold particles on the W-61 cell surface, whereas no immunogold particles were present on the cell surface of the Δxyn5 mutant (Fig. (Fig.1).1). The predicted amino acid sequence of Xyn5 also suggested that the hydrophobic SLH domain at its C terminus would interact with the cell surface (16). The cell surface binding effect of the SLH domain was assessed with the Δxyn5 mutant transformed by the plasmid carrying the gene for Xyn5 lacking the SLH domain (Xyn5ΔSLH), named pX5003. No immunogold particles were observed on the cell surface of the mutant harboring pX5003, while Xyn5ΔSLH was secreted into the culture supernatant (data not shown). Thus, we determined that Xyn5 binds on the cell surface and that the SLH domain of Xyn5 is required for interaction between Xyn5 and the cell surface.
In Gram-positive bacteria, the SLH domain is known as a structure that interacts with cell wall components, such as SCWPs and/or peptidoglycan itself (33, 34). To ascertain whether the SLH domain of Xyn5 interacts with these cell wall components of W-61, a GST-fused SLH domain polypeptide (GST-SLH) of Xyn5 was prepared, and the interaction between GST-SLH and purified W-61 peptidoglycan or SCWPs fractions was investigated. Under the assay conditions, GST-SLH binding with the insoluble peptidoglycan or SCWP molecules was coprecipitated by centrifugation while free GST-SLH remained in the supernatant. As shown in Fig. Fig.2,2, approximately 80% of the GST-SLH added was associated with the insoluble precipitate, indicating that a substantial fraction of GST-SLH bound to both SCWPs and peptidoglycan. In contrast, more than 90% of GST-SLH remained in the supernatant when it was incubated with peptidoglycan from E. coli DH5α. These results indicate that the SLH domain of Xyn5 specifically binds to both SCWPs and peptidoglycan of Paenibacillus sp. W-61.
Paenibacillus sp. W-61 formed a colony on the agar plate containing oat spelt xylan as a sole carbon source (42). In contrast, the Δxyn5 mutant (PW5001) was unable to form a colony on the plates containing oat spelt xylan or the water-insoluble fraction from oat spelt xylan, whereas it grew normally on the water-soluble xylan plate (data not shown). The growth of the Δxyn5 mutant was also restricted considerably in the water-insoluble xylan liquid medium (Fig. (Fig.3A).3A). The growth defect of the Δxyn5 mutant in the water-insoluble xylan medium was due to the deletion of the xyn5 gene and was not a polar effect of the cat insertion on downstream genes (data not shown). This growth defect in the Δxyn5 mutant was reversed in both xylan media by transformation of the plasmid pX5002, which carried the gene for intact Xyn5 (Fig. (Fig.3A).3A). The Δxyn5 mutant strain (PW5001) harboring pX5003 secreted Xyn5ΔSLH into the culture, but its ability to grow on water-insoluble xylan medium could not be completely recovered although it showed almost the same growth as the wild type in water-soluble xylan medium (Fig. (Fig.3A).3A). These results indicated that cell surface-bound Xyn5 is indispensable for water-insoluble xylan utilization, whereas an excess amount of secreted Xyn5ΔSLH could partially substitute for the function of cell-bound Xyn5.
There are two possible reasons for the Δxyn5 mutant's deficit in water-insoluble xylan utilization: (i) purified Xyn1 and Xyn3 rarely use water-insoluble xylan as a substrate, and therefore these xylanases contribute minimally to water-insoluble xylan degradation (26, 43); or (ii) the mutant could not produce these xylanases. The zymography experiment clearly indicated that five xylanases were present while the W-61 culture was growing in water-insoluble xylan medium (Fig. (Fig.3B).3B). Trace amounts of Xyn1 and Xyn3 were detected by Western immunoblotting in the Δxyn5 mutant when cells were cultured in water-insoluble xylan medium (data not shown). The total xylanase activity when water-soluble xylan was used as a substrate in the cultures of the Δxyn5 mutant and W-61 cells was determined to be 43 and 221 U/cell, respectively. Additionally, the amounts of xyn1 and xyn3 transcripts from Δxyn5 mutant cells growing in the water-insoluble xylan medium were at 10% of the level of the transcripts in the W-61 strain (Fig. (Fig.4A).4A). These results indicate that Xyn1 and Xyn3 expressed at basal levels by the Δxyn5 mutant growing in the water-insoluble xylan medium are not sufficient for water-insoluble xylan utilization by the mutant.
The Δxyn5 mutant actively expressed three xylanases, Xyn1, Xyn2, and Xyn3, in the water-soluble xylan medium (Fig. (Fig.3C),3C), and the total xylanase activities were 2-fold higher than those in water-insoluble xylan medium (80 U/cell). This would imply that water-soluble xylan medium contains possible signal molecules for xyn1 and xyn3 expression in the Δxyn5 mutant and that the existence of cell surface Xyn5 is associated with Xyn1 and Xyn3 production in water-insoluble xylan medium. The Δxyn5 mutant complemented by the intact xyn5 allele (PW5001/pX5002) restored the total xylanase activity to 54% (120 U/cell) of wild-type cells (Fig. (Fig.3B).3B). The value was almost identical to total xylanase activities (Xyn1 to Xyn5) secreted by the wild-type W-61 strain in the water-soluble xylan medium (Fig. (Fig.3C3C).
As described above, the expression of xyn1 and xyn3 was limited in the Δxyn5 mutant (PW5001) when the cells were grown in water-insoluble xylan medium (Fig. (Fig.3B),3B), yet the mutant was able to synthesize Xyn1 and Xyn3 when it was grown in water-soluble xylan medium (Fig. (Fig.3C).3C). The difference between water-insoluble xylan and water-soluble xylan was the degree of polymerization of xylose units but not the degree of substitution of arabinose and uronic acid for the main chain (data not shown). Moreover, water-soluble xylan contains more soluble xylo-oligosaccharides than water-insoluble xylan (data not shown). Thus, we hypothesized that xylo-oligosaccharides, which hydrolyzed water-soluble xylan by Xyn1 and Xyn3 (26, 43), would serve as signal molecule(s) for expression of xyn1 and xyn3 genes and also that Xyn5 would release these molecules digested with water-insoluble xylan. To assess this hypothesis, the induction of xyn1 and xyn3 expression in wild-type and Δxyn5 mutant cells was tested by using xylo-oligosaccharides prepared from the water-insoluble xylan hydrolysate due to Xyn5 and a commercial xylo-oligosaccharide mixture with a degree of polymerization ranging from 2 and 7. In the presence of both xylo-oligosaccharide preparations, the Δxyn5 mutant produced both Xyn1 and Xyn3 in amounts comparable to those produced by wild-type W-61 (Fig. (Fig.5A).5A). The Δxyn5 mutant commenced xyn1 and xyn3 transcription within 3 h after being transferred from glucose medium into xylo-oligosaccharide medium (Fig. (Fig.5B).5B). These results suggest that the xylo-oligosaccharides act as inducers for xyn1 and xyn3 expression. To identify the inducer molecule(s) for the expression of xyn1 and xyn3 existing in the xylo-oligosaccharide mixture, the induction ability of individual xylo-oligosaccharide molecules separated by high-performance liquid chromatography was tested. As shown in Fig. Fig.5C,5C, excess amounts (10 μM) of xylobiose and xylotriose in medium I led to the expression of xyn1 and xyn3 in the Δxyn5 mutant while trace amounts of the transcripts could be detected when the concentration of xylose was in excess. Xylotetraose, xylopentaose, and xylohexaose resulted in a 2- to 3-fold decrease in the number of transcripts (Fig. (Fig.5C).5C). These xylo-oligosaccharides can be hydrolyzed by the basally expressed Xyn1 to xylobiose and xylotriose, which may act as inducers for xylanase genes. Xyn5 hydrolyzed the water-insoluble xylan and liberated xylo-oligosaccharides (30) that were once again able to act as inducers of xylanase genes.
Given that the Xyn5 hydrolysate of water-insoluble xylan is required for expression of the xyn1 and xyn3 genes, xyn5 is expressed prior to xyn1 and xyn3 induction (Fig. (Fig.4B).4B). W-61 cell growth reached the stationary phase after 12 h (data not shown), but transcription of xyn5 was detected at 3 h, during the early exponential phase, and it reached maximum levels at 4.5 h. These levels of transcription were maintained for 3 h and then rapidly declined between 7.5 and 9 h of the late exponential growth phase. Western blotting revealed that the Xyn5 protein was first detected at 5 h and increased in conjunction with cell growth from 6 to 9 h before attaining a plateau (Fig. (Fig.6).6). Expression of xyn1 was observed from 4.5 to 6 h (Fig. (Fig.4B),4B), and Xyn1 could be detected at 6 h, but after this time, the amount of Xyn1 sharply increased, reaching almost maximal levels after 10 h (Fig. (Fig.6).6). The expression of xyn3 was detected at 6 to 7.5 h, and synthesis of Xyn3 occurred from 8 to 10 h (Fig. (Fig.4B4B and and6).6). These results indicate that synthesis of Xyn5 preceded that of Xyn1 and Xyn3 during growth of the W-61 strain after the cells were transferred to the medium with water-insoluble xylan as a sole carbon source. That induction of Xyn5 preceded expression of Xyn1 and Xyn3 during exponential growth of W-61 on water-insoluble xylan medium also suggests that Xyn5 released the inducer molecule(s) for xyn1 and xyn3 from water-insoluble xylan.
Since xyn1 and xyn3 were induced by xylobiose and xylotriose, we postulated that the xyn5 promoter would also be regulated by xylo-oligosaccharides. To examine this, we constructed pX5005, which has the xyn5 promoter fused to a luciferase reporter gene. We found that expression of xyn5 was induced during the early log phase in W-61 cells (Fig. (Fig.4B).4B). Analysis of W-61 cells harboring pX5005 revealed that 10 μM xylotriose significantly enhanced expression of the xyn5 promoter at 4 h, whereas limited induction was observed using xylobiose under the same conditions (Fig. (Fig.7).7). These results suggest that xylotriose is the minimum chemical unit sufficient to induce expression of xyn5. Xylotetraose, xylopentaose, and xylohexaose demonstrated almost equivalent luminescence, but this may be due to the effects of their hydrolysis by xylanases present in the culture (Fig. (Fig.7).7). The enhancement of xyn5 expression was observed by Western blotting in the presence of at least 0.1 μM xylotriose. Induction of xyn1 and xyn3 required more than 1 μM xylobiose and xylotriose (data not shown). We propose that Xyn5 on the cell surface plays a major role in water-insoluble xylan utilization by generating xylo-oligosaccharides as inducers not only for xyn1 and xyn3 but also for xyn5.
Since xylan is too large to pass through the bacterial cell membrane, the small amounts of soluble low-molecular-weight catabolite released from polymeric compounds by constitutive xylanases are used as the signal molecules to accelerate synthesis of the respective xylanases (31). The regulation of the synthesis of both cellulase and xylanase in fungi has been well studied, and a number of low-molecular-weight inducer molecules promoting endoglucanase formation have been identified (38). The inducer xylobiose stimulated only the synthesis of xylanase in resting cells of Trichoderma reesei, whereas sophorose induced the formation of both cellulase and xylanase. Analysis of the sophorose-induced enzyme system revealed that most of the xylanase activity could be attributed to a nonspecific endoglucanase while specific xylanases were induced in relatively small amounts (38).
The xylanolytic bacteria, Paenibacillus sp. W-61, secretes five xylanases from different GH families, and the activities of these xylanases are increased at different stages of growth on water-insoluble xylan medium. Xyn5 plays an important role in water-insoluble xylan utilization among these xylanases because Xyn5 exhibits 3.5-fold and 5.4-fold higher hydrolytic activity in water-insoluble xylan than Xyn3 (8, 30) and Xyn1, respectively. Then, the state of Xyn5 localized on the cell surface involves not only the efficient hydrolysis of water-insoluble xylan but also the induction of extracellular xylanase genes, xyn1, xyn3, and xyn5. According to this water-insoluble xylan utilization system, Xyn5 on the cell surface acts as a sensor for the substrate, even in cells grown in medium without water-insoluble xylan. Basal levels (10 ng) of Xyn5 were observed by Western blotting in 2 × 107 cells grown in medium containing glucose as the sole carbon source. After cells were transferred into water-insoluble xylan medium, the cell surface Xyn5 of W-61 hydrolyzed water-insoluble xylan to liberate xylobiose and xylotriose, and the resultant xylotriose enhanced xyn5 gene expression at logarithmic growth phase. Indeed, xylobiose and xylotriose were found in the degradation products when water-insoluble xylan was incubated with mutant cells producing only cell surface-localized Xyn5 (Δxyn1 and Δxyn3) (data not shown). Oat spelt xylan is classified as arabinoxylan (40), and commercially purchased product contains, relative to the total sugar content, 79% xylose, 6% arabinose, 12% glucose (that would come from remaining cellulose), and 3% uronic acid. The sugar content of the water-insoluble fraction of oat spelt xylan was similar to that of oat spelt xylan, whereas the sugar content of the water-soluble fraction consisted of 57% xylose, 15% arabinose, 25% glucose, and 3% uronic acid (data not shown). The average degree of polymerization of water-soluble xylan was estimated as ca. 100 to 300, which was much less than that of water-insoluble xylan (ca. 1,000 to 2,000). These facts indicate that more than half of xylose units in the main chains of xylan in both water-soluble and water-insoluble fractions were not substituted by arabinose and uronic acid and that Xyn5 released the same hydrolytic products (such as xylobiose and xylotriose) from both xylan fractions (data not shown). Furthermore, Xyn5 generated the same hydrolytic products from commercial glucurono- and arabinoxylans as well as from the oat spelt xylan, and the growth of Δxyn5 mutant was also restricted in water-insoluble fractions of these substrates (data not shown). Thus, Xyn5 generates xylo-oligosaccharides from the xylan main chain and is not influenced by substitutions. The resulting xylotriose was used as the activator of xyn5 to enhance cell-bound Xyn5, followed by xyn1 and xyn3 expression induced by xylobiose and xylotriose. The genes xyn1 and xyn3 required more than 10 times the extracellular concentration of xylo-oligosaccharide inducers than xyn5. The deficiency in growth of the mutant expressing soluble Xyn5 (Xyn5ΔSLH) may be explained by the focal concentration of inducer molecules around the cells. Surface-bound Xyn5 should have served to sustain the higher inducer concentration in the environment around the cell even if basal amounts of inducer molecules existed, whereas the inducer molecules generated by soluble enzymes easily diffuse into the medium so that Xyn5 did not reach a sufficient concentration.
We identified inducers of the water-insoluble xylan degradation system in Paenibacillus sp. W-61, but the details of this induction mechanism remain to be elucidated. In Bacillus subtilis, a representative Gram-positive bacterium, three components, namely, the cis-acting catabolite responsive element (cre), catabolite control proteins (CcpA and CcpB), and the heat-stable protein of the phosphotransferase system (Hpr), have been identified as being involved in catabolite repression (39). A GH family 11 xylanase gene, xynA, of Bacillus stearothermophilus No. 236 is known to be susceptible to glucose effects, with a potential cre sequence found upstream of this gene (3). We found typical cre sequences (nt −53 to 38) within the coding region of xyn5 genes and suggest that they take part in repression control of xylanase gene systems, whereas xyn1 and xyn3 do not contain homologous cre sequences. The differences in inducer concentration and nucleotide sequences of the 5′ flanking region suggest that xyn5 has a different regulation system from xyn1 and xyn3.
The SLH domain of Xyn5 is required for its interaction with the cell surface. The CBM9 and CBM22 domains of Xyn5, which bind the bacterial cell to cellulose (16), are required for the efficient hydrolysis and utilization of water-insoluble xylan. SLH domains, which bind very strongly to the cell wall components via noncovalent interactions, were originally discovered in glycoproteins forming the regular surface layer observed in some prokaryotic organisms, such as archaea and bacteria (33). SCWPs have been identified as cell wall polymers involved in the anchoring of surface layer proteins to the bacterial cell surface (15, 23). We determined that the SLH domain of Xyn5 is capable of binding to both peptidoglycan and SCWPs (Fig. (Fig.2).2). The molecular basis of affinity and specificity of interaction between the SLH domain and cell wall polymers remain to be resolved, and the SLH domain of Xyn5 may become one of the candidates to investigate the interaction between the SLH domain and cell wall components. SLH domains are found in other proteins displayed on the cell surface, namely, polysaccharide degradation enzymes such as cellulases, amylases, hemicellulases, galacturonidases, keratanases, or pullunases (5). We have proposed that cell surface-bound Xyn5 plays two important roles in water-insoluble xylan utilization: (i) constitutively expressed Xyn5 releases xylo-oligosaccharide inducers for all the xyn genes including xyn5 from water-insoluble xylan, and (ii) anchoring of cells to the substrate via Xyn5 enables greater accessibility to the degraded products of water-insoluble xylan. As a result the cells acquire an efficient ability to utilize water-insoluble xylan rather than only producing secreted xylanases. The activation of a xylan degradation system by an inducer generated by a cell surface-bound xylanase is a general principle among xylanolytic bacterium. Several insoluble glycoside hydrolytic bacteria are known to have the Xyn5-like cell surface-localized hydrolytic enzymes. Seven species of xylanolytic bacteria produce both multiple soluble xylanases and Xyn5-like cell wall-anchored xylanases having a multidomain structure. This consists of a central glycoside hydrolase family 10 catalytic domain flanked on the N-terminal side by CBM22 and on the C-terminal side by CBM9 (37). These xylanolytic bacteria include Paenibacillus sp. JDR-2 (37), Thermoaerobacterium saccharolyticum (20), Thermoanaerobacterium sp. strain JW/SL-YS 485 (22), and Caldicellulosiruptor sp. strain Rt69B.1, and the domain structure of the cell wall-anchored xylanases have been characterized (25). Additionally, the xylanases present in Paenibacillus curdianolyticus B-6 (accession number ABZ80916.1), C. thermocellum ATCC 27405 (AAA23227.1), and Thermoaerobacterium themosulfurigenes (AAB08046) have the Xyn5-like domain structure; however, the function of these cell surface-localized xylanases and their role in the xylan-degrading enzyme system of xylanolytic bacteria are yet to be clarified. These cell-anchored xylanolytic enzymes may have similar functions to Xyn5, and it is possible that the existence of cell surface-bound enzymes in such bacteria protects cells from the degraded products of insoluble substrates around them and also has a role in the strategy for the uptake of soluble substrates to the cells.
We are grateful to T. Satoh and K. Itoh for their skillful technical help in the operation of the electron microscope. We thank Y. Kato (Hirosaki University) for his useful suggestions and his large stock of knowledge about the structure of xylan.
This study was supported in part by the Hayashi Memorial Foundation for Female Natural Scientist of Japan (to M.F.).
Published ahead of print on 12 February 2010.