| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 Institute of Biology Leiden, Leiden University, Fungal Genetics Research Group, Wassenaarseweg 64, 2333 AL Leiden, The Netherlands
2 Department of Microbiology, Groningen Biomolecular Sciences and Biotechnology Institute (GBB), University of Groningen, Kerklaan 30, 9751 NN Haren, The Netherlands
3 Centre for Carbohydrate Bioprocessing TNO-University of Groningen, Kerklaan 30, 9751 NN Haren, The Netherlands
4 Microbiology, Fungal Genomics Group, Wageningen University, Dreijenlaan 2, 6703 HA Wageningen, The Netherlands
5 TNO Quality of Life, Business Unit Innovative Ingredients and Products, Rouaanstraat 27, 9723 CC Groningen, The Netherlands
6 TNO Quality of Life, Business Unit Microbiology, Utrechtseweg 48, 3500 AJ Zeist, The Netherlands
Correspondence
Arthur F. J. Ram
ram{at}rulbim.leidenuniv.nl
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
creA mutant strain of A. niger revealed that expression of the extracellular inulinolytic enzymes is under control of the catabolite repressor CreA. Expression of the inulinolytic enzymes was not induced by fructose, not even in the creA background, indicating that fructose did not act as an inducer. Evidence is provided that sucrose, or a sucrose-derived intermediate, but not fructose, acts as an inducer for the expression of inulinolytic genes in A. niger.
A supplementary table and figure are available with the online version of this paper.
These authors contributed equally to this work.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
-2,1-glycosidic bonds, and usually followed by a terminal glucose moiety. Inulin is present as storage polysaccharide in roots and tubers of plants such as Jerusalem artichoke, chicory and dahlia (Cairns, 2003
). Its presence has also been implicated in protection against water deficit in dry and cold conditions (Hendry & Wallace, 1993; Pilon-Smits et al., 1995). Inulin in plants is synthesized by the concerted action of two fructosyltransferases, with sucrose as the primary fructosyl donor (see for review Ritsema & Smeekens, 2003). Inulin has attracted considerable research attention because it is a relatively inexpensive and abundant substrate for the production of fructose-rich syrups, as well as a source for the production of fructo-oligosaccharides (FOS). Both fructose syrups and FOS are regarded as ‘functional foods' since they positively influence the composition of the intestinal microflora (Yun, 1996; Roberfroid & Delzenne, 1998; Kaplan & Hutkins, 2003).
Yeasts and filamentous fungi employ various enzymes to degrade inulin and sucrose (Pandey et al., 1999). Apart from displaying substrate hydrolysis, some of these enzymes can also perform transfructosylation reactions, producing the trisaccharide 1-kestose from sucrose (Rehm et al., 1998; Sangeetha et al., 2004; Yanai et al., 2001) and even longer fructo-oligosaccharides (Heyer & Wendenburg, 2001). Currently, all known fungal inulin-modifying enzymes are grouped together in family 32 of glycoside hydrolases (GH32) (http://afmb.cnrs-mrs.fr/CAZY/index.html) (Coutinho & Henrissat, 1999). Members of family GH32 share conserved amino acid motifs and possess a similar three-dimensional protein structure (Pons et al., 1998; Alberto et al., 2004; Nagem et al., 2004).
Aspergillus niger degrades inulin using both endo-inulinases (EC 3.2.1.7), encoded by the inuA and inuB genes (Ohta et al., 1998; Akimoto et al., 1999), and an exo-inulinase (EC 3.2.1.80), encoded by the inuE gene (Moriyama et al., 2003). Endo-inulinase hydrolyses inulin internally to produce mainly inulotriose and -tetraose (Akimoto et al., 1999), whereas exo-inulinase hydrolyses the terminal -2,1-fructosidic bonds in both sucrose and inulin (Arand et al., 2002; Kulminskaya et al., 2003; Moriyama et al., 2003). Invertase (-fructofuranosidase, EC 3.2.1.26), encoded by the suc1 gene (Boddy et al., 1993), hydrolyses the -2,1-glycosidic bond in sucrose to produce fructose and glucose (L'Hocine et al., 2000). A specific -fructosyltransferase activity (EC 2.4.1.9) without significant invertase activity has been purified from A. niger strain AS0023. This enzyme transfers fructose residues from the non-reducing terminal -2,1-glycosidic bond in sucrose to another sucrose or inulin molecule to form kestose or higher fructo-oligosaccharides (L'Hocine et al., 2000). Unfortunately, the gene encoding this enzyme activity has not been identified and characterized yet.
Recent advances in the genome sequencing of A. niger opened possibilities to further exploit this fungus to identify additional inulin-modifying enzymes. The full genomic sequence of A. niger was made available to us by DSM Food Specialties (http://www.dsm.com). Based on deduced amino acid similarities, we have identified six putative proteins that belong to family GH32. Apart from the three known fungal enzymes (InuA/B, InuE and Suc1), three new putative inulin-modifying enzymes were identified. The coding sequence for one of them appears to be a pseudogene (inuQ), while the other two genes encode intracellular invertase-like proteins that were named SucB and SucC. The transcriptional regulation of these five putative inulin/sucrose-modifying proteins in relation to various carbon sources has been studied in further detail.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
). The A. niger strain used for the sequencing of the genome by DSM is CBS 513.88 (a natural derivative of strain NRRL 3122). Strain AB4.1 is a pyrG derivative of N402 (van Hartingsveldt et al., 1987) and was used to construct the creA deletion strain. A. niger strains were grown in minimal medium (MM) (Bennett & Lasure, 1991) containing 7 mM KCl, 11 mM KH2PO4, 70 mM NaNO3, 2 mM MgSO4, 76 nM ZnSO4, 178 nM H3BO3, 25 nM MnCl2, 18 nM FeSO4, 7.1 nM CoCl2, 6.4 nM CuSO4, 6.2 nM Na2MoO4 and 174 nM EDTA. Erlenmeyer flasks of 300 ml were inoculated with 2x106 spores ml–1 and incubated at 30 °C in a rotary shaker at 300 r.p.m. for 21 h. Each flask contained 100 ml MM (pH 6.5) supplemented with 0.1 % (w/v) Casamino acids and 2 % (w/v) carbon source. Glucose, sucrose (BDH Chemicals), xylose, fructose and maltose (Sigma-Aldrich), inulin (Sensus Frutafit, Cosun) and starch (Windmill Starch, Avebe) were used as carbon sources. For transfer experiments, N402 was pregrown in MM supplemented with 2 % (w/v) xylose or 2 % (v/v) glycerol and 0.1 % (w/v) Casamino acids for 18 h at 30 °C on a rotary shaker at 300 r.p.m. Mycelium was harvested by suction over a nylon membrane and washed with MM without carbon source. Aliquots of 1.5 g wet weight of mycelium were transferred to 70 ml MM containing 1 % (w/v) of various carbon sources and incubated at 30 °C with agitation. Mycelial samples were taken at different time points by harvesting over a Miracloth filter followed by freezing in liquid nitrogen. The samples were stored at –80 °C prior to the isolation of total RNA. Conidiospores were obtained by harvesting spores from a plate of complete medium (minimal medium with 0.5 %, w/v, yeast extract and 0.1 %, w/v, Casamino acids) containing 1 % (w/v) glucose, after 4–6 days of growth at 30 °C, using a 0.9 % (w/v) NaCl solution.
Transformation of A. niger AB4.1 was as described by Punt & van den Hondel (1992), using lysing enzymes (L1412, Sigma-Aldrich) for protoplast formation. The bacterial strain used for transformation and amplification of recombinant DNA was Escherichia coli XL-1 Blue (Stratagene). Transformation of XL-1 Blue was performed by the heat-shock protocol as described by Inoue et al. (1990).
Database mining of A. niger genome.
The A. niger CBS513.88 genome has been determined by random sequencing of selected BACs to a 7.5-fold coverage. The resulting genome sequence (35.9 Mb) consists of approximately 400 contigs, which are assembled into 19 supercontigs (Dr N. van Peij, DSM, personal communication). The sequence of the A. niger genome is currently available to academic groups and non-profit organizations on request (hans.roubos{at}dsm.com) after signing a Material Transfer Agreement and will be generally available after publication of the A. niger genome sequence (H. Pel and others, in preparation). Accession numbers of currently described members of families GH32 and GH68 were selected from the Carbohydrate-Active Enzymes server at http://afmb.cnrs-mrs.fr/CAZY/ Coutinho & Henrissat, 1999), and the corresponding protein sequences were extracted from the GenBank/GenPept database and Swiss-Prot database released at http://www.ncbi.nlm.nih.gov/entrez/ and http://www.expasy.org/sprot/. Sequences were aligned with the CLUSTALW program (Thompson et al., 1994; Chenna et al., 2003) and transformed in a hidden Markov model (HMM) profile (Eddy, 1998) with the HMMbuild program from the HMMer package at http://hmmer.wustl.edu/. Subsequently the A. niger genome was searched using the HMM profiles and the Wise2 package from http://www.ebi.ac.uk/Wise2/. Multiple sequence alignment of known fungal fructan-modifying enzymes, based on full-length predicted protein sequences (Table 1), was performed using the CLUSTALW interface in MEGA 3.1 (www.megasoftware.com) with gap-opening and extension penalties of 10 and 0.2, respectively. Bootstrap test of phylogeny was performed by the neighbour-joining method using 1000 replicates.
|
-32P]dCTP-labelled probes were synthesized using the Rediprime II DNA labelling system (Amersham Pharmacia Biotech) according to the instructions of the manufacturer.
Disruption of the carbon catabolite repressor CreA in A. niger.
The plasmid used to disrupt the creA gene was constructed as follows. The DNA fragments flanking the creA ORF were amplified by PCR using N402 genomic DNA as template: 1.4 kb of 5' flanking DNA and 0.9 kb of 3' flanking DNA was amplified by PCR using primers CreAP1f and CreAP2r, CreAP3f and CreAP4r (supplementary Table S1), respectively. Each primer was adapted with a restriction site for further cloning. The amplified PCR fragments were digested with NotI and BamHI or BamHI and KpnI respectively, and cloned into pBlueScriptII SK to obtain plasmids pF5 and pF3. Subsequently, pF3 was digested with BamHI and KpnI, and the fragment obtained was ligated into BamHI- and KpnI-digested pF5 to give pF53. pF53 was digested with SalI and BamHI and inserted with the SalI–BamHI fragment containing the Aspergillus oryzae pyrG gene, obtained from plasmid pAO4-13 (de Ruiter-Jacobs et al., 1989), resulting in the creA disruption plasmid pXY1.1. Plasmid pXY1.1 was linearized with NotI and transformed into AB4.1. Uridine-prototrophic transformants were selected by incubating protoplasts on agar plates containing MM without uridine. Transformants were purified and genomic DNA was isolated and analysed by PCR to identify possible creA strains. Primer pairs used to identify homologous recombination of the creA deletion construct on the creA locus were CreAP5f and PAO10 or PAO9 and CreA6f. Primer pairs used in the PCR to analyse the presence of the wild-type creA gene were CreAP5f and CreAP7r, CreAP8f and CreAP6r (supplementary Table S1). Three independent creA deletion strains with identical phenotypes were obtained and designated XY1.1, XY1.2 and XY1.3. Strain XY1.1 was further used for analysis of the expression of inulin-modifying enzymes and we will refer to this strain as the creA strain in the remainder of the paper. For complementation of the creA strain, the creA gene, including 1.3 kb promoter and 0.8 kb of terminator sequences, was amplified by PCR using primers CreAP1f and CreAP6r. The PCR product of 3.5 kb was cloned into pGEMT-easy (Promega) and co-transformed with pAN7.1 (Punt et al., 1987) to creA strain XY1.1 to generate XY1.1-CreA.
Nucleotide accession numbers.
The A. niger CBS513.88 DNA sequences encoding family GH32 members, including 1000 bp up- and downstream of the ORF, and their predicted protein sequences were obtained from DSM (Dr G. Groot). The sequence data have been submitted to the GenBank database under accession numbers DQ233218 (sucA), DQ233219 (sucB), DQ233220 (sucC), DQ233221 (inuA), DQ233222 (inuE) and DQ233223 (inuQ).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
; Pons et al., 1998). The two families are structurally similar (clan GH-J), sharing a similar fivefold -propeller fold (Meng & Futterer, 2003; Nagem et al., 2004). Protein sequences from members of families GH32 and GH68 were extracted from the GenBank/GenPept and Swiss-Prot databases and were used to construct HMM profiles to identify additional members in the genome of A. niger CBS513.88. Family GH68 profiles did not return any significant matches. The family GH32 profile returned five significant sequences. In addition to three family GH32 members already described for A. niger – invertase (Suc1, Boddy et al., 1993), exo-inulinase (InuE, Moriyama et al., 2003) and endo-inulinases (InuA/InuB, Ohta et al., 1998) – two new members were identified, which were named SucB and SucC.
A neighbour-joining tree based on phylogenetic analysis of all currently available functionally described family GH32 fungal protein sequences was constructed. As shown in Fig. 1, four subgroups of fungal inulinolytic enzymes can be distinguished: (i) exo-inulinases/fructosyltransferases, (ii) endo-inulinases, (iii) yeast invertases/inulinases and (iv) invertases from filamentous fungi. The complete inventory and the notion that these four groups are evident within the fungal GH32 enzymes had not been noticed in earlier studies in which phylogenetic trees of GH32 family proteins had been constructed (Pons et al., 1998; Moriyama et al., 2003). In the secondary structure of microbial GH32 proteins eight well-conserved domains (A, B, B1, C, D, E, F and G, respectively) can be distinguished (Pons et al., 1998; Ohta et al., 1998). Domains A, D and E (designated blocks I, II and III by Pons et al., 2004) contain highly conserved acidic residues that are located in the active site of members of family GH32. These highly conserved acidic domains, as well as the other conserved domains, are also present in most of the fungal fructan-modifying enzymes (Fig. 2).
|
|
). In the genome sequence of A. niger strain CBS513.88, a single gene encoding an endo-inulinase could be identified. The presence of single-copy endo-inulinase genes in A. niger strains CBS513.88 and N402 was confirmed by Southern blot analysis (data not shown). Sequence comparison of the endo-inulinases of strains CBS513.88 and 12 revealed that the enzyme from strain CBS513.88 displayed higher similarity to InuA (9 amino acid differences) than to InuB (15 amino acid differences) of strain 12. The single endo-inulinase in strain CBS513.88 has been designated InuA.
Besides the three known inulinolytic activities, two previously unknown GH32 family members were identified in the A. niger CBS513.88 genome sequence. Unlike the SucA, InuA and InuE proteins, SucB and SucC lack an N-terminal signal sequence, or any other recognizable targeting signal, which suggests that the proteins are localized intracellularly, in the cytosol. The SucB and SucC protein sequences were analysed using SecretomeP 1.0b prediction (http://www.cbs.dtu.dk/services/SecretomeP-1.0) (Bendtsen et al., 2004) to assess if these enzymes might be secreted via a non-classical secretion pathway. Both enzymes have an NN-score close to the threshold value of 0.6 (0.651 for SucB and 0.586 for SucC), which does not exclude the possibility that the two proteins are secreted via a non-classical secretion pathway. This result should be interpreted with care, as the program has been trained using sequences of human non-classical exported proteins.
Phylogenetic analysis indicated that SucB and SucC group together with fungal invertases (Fig. 1
). Comparison of the deduced amino acid sequence of SucB with all functionally described GH32 family proteins revealed highest identity to the A. niger SucA protein (24 % identity, 41 % positives, e-value of 4x10–32). SucC also displays the highest identity to the A. niger SucA protein (28 % identity, 42 % positives, e-value of 2x10–43). Pairwise comparison indicated that SucB and SucC have higher identity to each other (35 % identity, 52 % positives, e-value of 2x10–97) than to any other functionally annotated GH32 family member (Fig. 1). The sucB and sucC genes encode proteins of 617 and 601 amino acids respectively, and contain all conserved domains (A–G), including the conserved acidic residues in domains A, D and E (Fig. 2).
Identification of glycoside hydrolase family 32 members in other fungal genomes
The genomes of the ascomycete fungi Aspergillus nidulans, Aspergillus fumigatus, Neurospora crassa, Gibberella zeae and Magnaporthe grisea and the basidiomycete fungus Ustilago maydis were analysed for the presence of GH32 family members. The results are summarized in Table 2 and a phylogenetic tree is available as supplementary Fig. S1. A. nidulans contains two genes that encode proteins belonging to the GH32 family. AN5012.2 displays the highest identity to the A. niger InuE and probably represents the extracellular A. nidulans sucrose-hydrolysing activity described by Vainstein & Peberdy (1990, 1991). An invertase similar to the A. niger SucA protein appears to be absent in A. nidulans. The second member of the GH32 family (AN3837.2) has high identity to the A. niger SucB protein (Table 2). As in A. niger, the protein is predicted to be intracellular in A. nidulans, as an N-terminal signal sequence is not present. Orthologues of an endo-inulinase (InuA-like) or a second intracellular invertase-like protein (SucC-like), as found in A. niger, were not found in the A. nidulans genome. A. fumigatus contains four genes that encode proteins belonging to the GH32 family. Based on the sequence alignments, Afu5g00530 and Afu5g00480 are likely to encode the endo- and exo-inulinases, respectively. Afu6g05000 is highly similar to the A. niger SucB protein and also predicted to be intracellularly localized. The fourth GH32 family member, Afu2g01240, shows the highest sequence identity to yeast-like invertases, also reflected in the neighbour-joining tree (Fig. S1). N. crassa contains only a single GH32 family member. This protein displays the highest sequence identity with -fructofuranosidases of bacterial origin (Bacillus megaterium FruA; e-value 3x10–72) (Chiou et al., 2002) and groups in the fungal tree together with the yeast-like invertases (Fig. S1). G. zeae (anamorph Fusarium graminearum) contains five GH32 family members. FG08415.1 is the putative orthologue of the A. niger SucA protein. In the original annotation, the protein lacks an N-terminal signal sequence. However, deleting the first 81 amino acids of the predicted protein sequence renders a protein of 619 amino acids, containing a predicted signal sequence. FG02339.1 is most homologous to the Saccharomyces cerevisiae Suc2 protein (e-value 3x10–82) and clusters together with yeast-like invertases. FG03288.1 encodes a protein with high identity to the SucC protein of A. niger. The protein is also predicted to be intracellularly localized. FG06451.1 shows the highest identity (e-value 1x10–72) to the Bacillus subtilus SacC protein, which has levanase activity (Martin et al., 1987). FG02067.1 is an interesting member of the GH32 gene family as it shows very limited sequence identity to the known enzymes. It is predicted to be an intracellular enzyme and shows the highest identity (e-value 2x10–09) to a cycloinulooligosaccharide fructanotransferase from Bacillus macerans (Kim & Choi, 2001). M. grisea contains five GH32 enzymes, four of which cluster in a separate branch different from the other fungal proteins (Fig. S1). One of the predicted enzymes (MG10748.4) is N-terminally truncated because the ORF is at the end of a contig. The annotation of MG10767.4 has been improved, resulting in a protein that contains all the conserved GH32 domains. MG02507.4 clusters in the group of yeast-like invertases (Table 2, Fig. S1), whereas the remaining four form a distinct group with high sequence identities to each other (e-value 1x10–112 or lower). Three of the enzymes are predicted to be secreted; for the fourth protein this is not known (N-terminally truncated because this ORF is at the end of a contig). It will be of interest to characterize the biochemical properties of those enzymes. Finally, U. maydis contains two GH32 family members. UM01945.1 encodes a protein that is most similar to yeast-like invertases, while UM03605.1 encodes a predicted intracellular protein most closely related to the A. niger SucC protein.
|
). With one exception (FG02067), these enzymes cluster as a distinct group in the phylogenetic tree (Fig. S1).
Transcriptional regulation of inulin-modifying enzymes in A. niger
The expression of the five putative inulin/sucrose-modifying enzymes identified in the genome of A. niger in relation to the presence of different carbon sources was studied by Northern blot analysis. RNA was extracted from A. niger N402 mycelia grown in minimal medium containing xylose, glucose, maltose, starch, fructose, sucrose or inulin (all 2 %, w/v) as sole carbon source. Expression of the inuE, sucA and inuA genes could be detected only on sucrose and inulin (Fig. 3a). sucB was not only expressed in the presence of sucrose and inulin; weak expression was also detected on other carbon sources (Fig. 3a). SucB expression on glucose and maltose was detected after longer exposure of the Northern blot (not shown). The detection of two different-sized mRNAs for the sucB gene suggests two different mRNA start sites, or two different polyadenylation sites. The presence or absence of one of the two mRNAs was not correlated with a particular carbon source. Expression of sucC was not detected on any of the carbon sources tested (results not shown). The differences in expression level of the various genes (relatively low on sucrose and high on inulin) might be caused by differences in the growth stage of the cultures. Growth of A. niger on sucrose is much faster than that on inulin. After 21 h, the sucrose-grown culture was in the mid-exponential phase of growth, while the inulin-grown culture was still in the early exponential phase (data not shown).
|
), this carbon source was chosen for pre-culturing. Mycelia were isolated before the transfer (t=0 h) and at specific time points after the transfer (2, 4, 8 and 24 h) and total RNA was isolated and subjected to Northern analysis. As expected, no expression of inulinolytic enzymes was observed during growth on xylose (Fig. 4, t=0 h). The expression of all four genes, inuE, sucA, inuA and sucB, was induced after transfer from xylose to sucrose and inulin (Fig. 4). No induction of any of the inulinolytic genes was observed on maltose, although a low level of expression of inuE could be observed. The induction of genes on sucrose was much faster and more dramatic compared to the response to inulin, reflecting the much faster growth on sucrose than on inulin (see above). The expression of the inuE, sucA and inuA genes decreased after 4 h growth on sucrose, probably due to the rapid utilization or conversion of the available sucrose. As inulin was utilized more slowly, expression levels remained higher for a longer period. The results in Fig. 4 also indicate that, although the different genes are co-regulated and all induced on sucrose or inulin, there were slight differences in expression pattern. Whereas the inuE transcript remains present most abundantly at 24 h after transfer to inulin, the levels of sucA and inuA mRNA had already decreased. We have no conclusive explanation for these observations; these differences may be due to differences in mRNA stability, or to expression also being under control of multiple regulatory mechanisms such as pH regulation, mediated by the PacC transcriptional regulator, or carbon catabolite repression, mediated by the CreA repressor protein. The presence of binding sites for PacC (5'-GCCARG-3') (see for review Penalva & Arst, 2002) and CreA (5'-SYGGRG-3') (see for review Ruijter & Visser, 1997) in the promoters of most of these genes might support the latter explanation.
|
creA background). The results shown in Fig. 3(a) indicate that the genes encoding the inulinolytic enzymes are not expressed on fructose in A. niger. This might be caused by the high concentration of fructose used, resulting in carbon catabolite repression mediated by CreA. Therefore, the expression of genes encoding inulinolytic enzymes was also examined in a creA mutant of A. niger strain N402. A creA null mutant was generated and verified as described in Methods. The phenotype of the creA strain was similar to that described for severe creA mutants in other A. niger strains and in A. nidulans (Ruijter & Visser, 1997; Shroff et al., 1997), including reduced radial growth and reduced conidiation (data not shown). Transformation of the creA strain with the wild-type creA gene fully complemented the reduced growth and reduced conidiation phenotype of the creA strain (data not shown).
The expression analysis of the inuE gene in the creA strain showed that the expression of this gene was higher than in the wild-type strain, indicating that it is under control of carbon metabolite repression. Expression of inuE was found in the creA strain after growth on glucose, and to some extent on maltose and starch. This is in contrast to the expression of inuE on xylose and fructose, which was undetectable in both the wild-type and the creA strain. Apparently, the expression of the inuE gene requires an activator or inducer molecule and is not expressed under derepressing (creA) conditions. Inspection of the 1 kb promoter sequence of the inuE gene revealed the presence of 13 putative CreA-binding sites that might be involved in mediating repression, but their functionality has not been studied. Analysis of the expression of the other genes encoding inulinolytic enzymes showed a different expression pattern in the creA strain. sucA and inuA, both expressed specifically on inulin and sucrose in the wild-type strain, were expressed at a lower level (sucA), or not detectable (inuA) in the creA strain (Fig. 3b). Expression of sucA and inuA was not detected after growth on xylose and fructose, similar to what was observed for inuE. In contrast to the expression profile of the inuE gene, no expression was detected of sucA and inuA after growth on starch, maltose and glucose. Thus, although CreA-binding sites are present in the 1 kb promoter regions of the sucA and the inuA genes (six and five sites, respectively), the expression of both genes does not seem to be directly controlled by CreA. Disruption of catabolite repression in the creA strain could lead to the inactivation of complex pathways, which might bring about decreased or total loss of expression. The expression pattern of sucB also suggests that this gene is, like inuE, under carbon catabolite repression control. sucB is not exclusively expressed on inulin and sucrose, but also on starch, xylose and fructose, and to a lesser extent on maltose and glucose (Figs 3 and 4). Expression of the sucB gene in the creA background showed two remarkable features: (i) expression of sucB in the creA strain was also detected on maltose and glucose, in addition to the other carbon sources, indicating that CreA mediated catabolite repression of sucB on maltose and glucose in the wild-type strain; (ii) expression of sucB in the creA background favours the transcription of the smaller-sized mRNA molecule. The 1 kb promoter region of sucB contains seven putative CreA-binding sites. Expression of the sucC gene was also not detected in the creA mutant (data not shown). Clearly, the different responses of inuE, sucB and sucA/inuA in a creA mutant background suggest the involvement of additional factors other than the presence of an inducer and repression via CreA. These might include environmental factors such as pH, nitrogen availability and temperature.
Sucrose acts as an inducer of the inulinolytic system in A. niger
Fructose has been shown to induce expression of inulinase in the yeast K. fragilis (Grootwassink & Hewitt, 1983). The expression analysis of the inulinolytic genes in the wild-type A. niger strain and the creA strain after growth on fructose did not result in detectable expression of any of the genes, indicating that fructose is not the inducing molecule for expression (Fig. 3).
Further evidence that fructose did not act as an inducer for the expression of the inulinolytic genes was obtained in a transfer experiment. Wild-type A. niger strain N402 was pre-grown in 2 % (w/v) glycerol minimal medium for 18 h and mycelium was transferred to minimal medium containing decreasing concentrations of fructose: 50 mM, 5 mM, 500 µM, 50 µM, 5 µM, 500 nM, 50 nM, 5 nM, and no carbon sources. As shown in Fig. 5(a), this gradual decrease in fructose did not result in expression of the different inulinolytic genes, not even after 4 h of growth. A similar transfer experiment was performed to medium containing sucrose in an identical concentration series. As expected, sucrose induced expression of the genes encoding the inulinolytic enzymes (Fig. 5b). The induction of inuE, sucA, inuA and sucB reached the highest level at 50 µM sucrose, indicating that some form of repression at high sucrose concentrations may exist, e.g. catabolite repression by released glucose from sucrose hydrolysis. The addition of 50 µM or 5 µM of glucose, 1-kestose or 1-nystose did not trigger induction of the inulinolytic system (data not shown).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
). The Suc1 protein from A. niger strain B60 has been described as a 566 amino acid protein; its protein sequence is identical to that of the SucA protein encoded by the genome of A. niger CBS513.88. The Suc1 protein groups in the phylogenetic tree with the previously described fructosyltransferase FopA (Yanai et al., 2001). FopA was originally published as an enzyme from A. niger strain ATCC 20611, but this strain has been reclassified as an Aspergillus japonicus strain (ATCC culture collection database; http://www.lgcpromochem-atcc.com/). Compared to the Suc1 protein sequence (Boddy et al., 1993), FopA has an extra C-terminal extension consisting of 38 amino acids (Yanai et al., 2001). We have re-examined the suc1 coding region and from that analysis we predict the existence of a 55 bp intron sequence 1 bp upstream of the termination codon. Taking the intron sequence into account, the Suc1 protein is predicted to become 39 amino acids longer than the sequence published by Boddy et al. (1993). To verify our prediction, the suc1 cDNA was amplified using RT-PCR and sequenced, which confirmed the presence of the intron (data not shown). We have renamed the suc1 gene as sucA, to meet the general nomenclature rules for A. niger. Although the A. niger SucA protein is 67 % identical to the FopA protein, and both proteins cluster in the same branch of the phylogenetic tree, the enzymic activities of the two proteins appear to be different. FopA displays a much higher fructosyltransferase activity than Suc1 from A. niger strain B60 (Yanai et al., 2001). SucA, which is identical to Suc1, is therefore likely to encode a fructofuranosidase lacking any detectable fructosyltransferase activity under the given assay conditions. A second fungal fructosyltransferase-encoding gene was identified in Aspergillus sydowi (sftA) (Heyer & Wendenburg, 2001), which clusters with FopA and SucA (Fig. 1). The sftA gene was only expressed in the conidia of A. sydowi, and its product SftA is capable of producing fructo-oligosaccharides up to 40 fructose units long (in vitro). High-molecular-mass polymers were detected when intact conidia were incubated with sucrose as substrate. Unexpectedly, these two transferases (SftA and FopA) did not show a higher sequence identity to each other than to SucA. Our sequence alignment did not reveal any obvious differences between the proteins that could explain their different reaction specificities. SucA orthologues were not identified in the genomes of A. nidulans and A. fumigatus.
A surprising finding was the presence of only one copy of an endo-inulinase gene in the A. niger genomes of strain CBS513.88 and N402. Previous studies with A. niger strain 12 revealed the presence of two very similar genes, inuA and inuB, both encoding endo-inulinase (Ohta et al., 1998). Expression analysis revealed that only the inuB gene is transcribed actively (Akimoto et al., 1999). Most likely, the presence of the two genes is a result of a recent duplication event that is specific for A. niger strain 12 and has not occurred in A. niger strains CBS513.88 and N402.
The inuE gene encodes a third known inulinolytic enzyme, an exo-inulinase, characterized from A. niger strain 12 (Moriyama et al., 2003). The putative exo-inulinase from CBS513.88 is 99 % identical to InuE from A. niger 12 and 100 % identical to Aspergillus foetidus fructosyltransferase 1-Sst (Rehm et al., 1998); the latter was shown to produce 1-kestose in the presence of high concentrations of sucrose. However, Moriyama et al. (2003) did not detect transfructosylation activity in the culture filtrate of a Pichia pastoris strain expressing inuE, although the enzyme was incubated at substrate concentrations where transfructosylation should occur (150 mM), as reported by Rehm et al. (1998). It appears unlikely that the three amino acid differences between 1-Sst and InuE affect reaction specificity since the differences involve similar amino acid residues: His199Gln, Gly476Ser and Thr499Ser. Moriyama et al. (2003) suggested that 1-Sst might be an exo-inulinase that possessed an additional fructosyltransferase activity in the presence of high concentrations of sucrose. Both proteins contain signal sequences at their N-termini and are secreted as extracellular proteins. inuE was specifically induced on sucrose and inulin and repressed on fructose and glucose (Moriyama et al., 2003). In view of the 100 % sequence identity of the A. niger CBS513.88 InuE and 1-Sst proteins, it is most likely that the A. niger InuE protein also has fructosyltransferase activity. However, further biochemical data about InuE are required. The InuE protein is also 91 % identical to the Aspergillus awamori exo-inulinase Inu1 protein (Arand et al., 2002). A. awamori Inu1 is the only enzyme in this branch of the phylogenetic tree with levanase (hydrolysis of -2,6 glycosidic linkages) activity. No transfructosylation activity has been reported for the A. awamori Inu1 protein. The different enzymic properties of the enzymes in this group are currently not well understood and need further biochemical investigation.
Two new invertase-like proteins (SucB and SucC) were identified in the genome of A. niger. As the two proteins contain all the conserved domains and the catalytic residues of GH32 family members it is very likely that these proteins contain sucrolytic or inulinolytic activities. Preliminary biochemical analysis of the recombinant SucB protein expressed in E. coli indeed indicates that SucB has hydrolysing activity on sucrose, 1-kestose and nystose as well as transfructosylation activity, resulting in the formation of 1-kestose and nystose from sucrose and 1-ketose, respectively (C. Goosen, unpublished results).
Both SucB and SucC are predicted to be intracellularly localized. The current gene models do not indicate the presence of typical hydrophobic signal sequences for targeting the protein into the endoplasmic reticulum in order to secrete the proteins via the secretory pathway. The algorithm (SecretomeP 1.0b prediction) used to predict the probability that SucB and SucC are exported via a non-classical secretion pathway was not conclusive, as the prediction scores of the proteins were close to the threshold values. Moreover, these values should be interpreted with care, as the program has been trained using sequences of human non-classical exported proteins as no non-classical protein export has been shown to be present in filamentous fungi. Also in the genomes of other filamentous fungi GH32 family members without a predicted N-terminal signal sequence were identified (Table 2), indicating that the presence of these intracellular enzymes is widespread among filamentous fungi and not specific to A. niger.
The presence of carbohydrate-degrading enzymes that do not contain a signal sequence is not limited to inulinolytic enzymes alone. In the genome of A. nidulans, 41 polysaccharide-degrading enzymes were predicted that lacked a signal sequence (de Vries et al., 2005). The presence of these intracellular enzymes strongly suggests that fungi are capable of transporting oligosaccharides into the cell which are subsequently hydrolysed by the intracellular enzymes. Alternatively, these intracellular enzymes may also possess transglycosylation activity, e.g. for the synthesis of inducer molecules that activate transcription factors.
The genome of A. niger also contained an ORF which showed homology to the group of exo-inulinases. However, to assemble this ORF encoding 137 amino acids, containing some of the conserved domains characteristic for family GH32 members, three putative frame shifts had to be corrected. Missing domains were not found in sequences adjacent to the predicted ORF. The genomic DNA region containing this putative inulinase (inuQ) was amplified by PCR from CBS513.88 and N402 and resequenced, which confirmed that the original DNA sequence was the correct one (data not shown). From this we concluded that inuQ is probably a pseudogene and not producing a functional protein. Northern analysis of mRNA isolated from cells grown on a variety of different carbon sources did not result in detection of inuQ mRNA, indicating that the gene is not transcribed under the conditions tested.
Expression analysis of the five genes revealed that the genes encoding the extracellular enzymes (SucA, InuE and InuA) are co-regulated and specifically expressed on sucrose and inulin. It is rather surprising that the inuA gene is induced by sucrose. Physiologically, there is no reason for the fungus to secrete this enzyme during growth with sucrose, since the enzyme does not hydrolyse the fructose-glucose disaccharide. However, a beneficial mechanism might have evolved, as sucrose and inulin might often be present together in plant material. Inulin is broken down primarily into fructose residues by the action of the exo-inulinase gene product and into inulo-oligosaccharides by the action of the endo-inulinase. The specific expression of sucA and inuE on inulin and sucrose is in agreement with previous observations (Wallis et al., 1997; Moriyama et al., 2003; Rehm et al., 1998). The expression of the single endo-inulinase (inuA) gene in A. niger N402 on sucrose and inulin is different from that described for inuA (no expression observed) and inuB (constitutively expressed on inulin, fructose and glucose) in A. niger strain 12 (Akimoto et al., 1999). Analysis of the promoter sequences of the inuA genes of the A. niger CBS513.88 and N402 strains revealed that they contain five putative CreA-binding motifs (SYGGRG) in their first 1000 bp. creA encodes a wide-domain regulatory protein that binds to the promoter of target genes to prevent or decrease expression if a favourable carbon source (such as glucose or fructose) is present (Dowzer & Kelly, 1991; Ruijter & Visser, 1997). No such binding sites were observed when analysing the upstream sequence region of the A. niger 12 inuB gene. This difference in the promoter sequences of inuA (A. niger CBS513.88 and N402) and of inuB (A. niger 12) may be responsible for their different expression patterns in relation to different carbon sources. Thus, A. niger strains CBS513.88, N402 and 12 differ both in the number of genes encoding endo-inulinases and in the way their expression is regulated.
Two lines of evidence led us to propose that sucrose, and not fructose, acts as an inducer for the expression of the genes encoding the inulinolytic enzymes. First, we showed that a low concentration of sucrose (50 µM initial concentration) induced the expression of the genes encoding the inulinolytic enzymes. The lower expression of the inulinolytic genes at higher sucrose concentrations suggested that sucrose (or the hydrolysis products of sucrose: glucose and fructose) caused carbon catabolite repression, possibly via the repressor protein CreA. The addition of 5 µM sucrose resulted in relatively low levels of expression. Lowering the sucrose concentration even further did not result in detectable mRNA levels of any of the genes after 4 h of growth. At this point we can not rule out the possibility that also these lower concentrations of sucrose might have induced expression of the inulinolytic enzymes at earlier time points. Assuming that the mRNAs might not be very stable, we might have missed the induction as we have analysed the expression only after 4 h. In an identical experimental set-up, also fructose was tested as an inducer, previously reported to act as an inducer for inulinase expression in the yeast K. lactis (Grootwassink & Hewitt, 1983). We obtained no evidence that fructose acted as an inducer for the expression of the inulinolytic enzymes in A. niger. Formally, it is possible that the inulinolytic genes are induced by low influx levels of fructose which escaped detection after 4 h of growth. The addition of low concentrations of fructose may have resulted in starvation and the inability of the fungus to induce expression due to a lack of energy. One could postulate that fructose can act as an inducer at low concentrations but repress expression at higher concentrations. If this is the case, growth of the creA strain on fructose should lead to a high level of expression of the genes encoding the inulinolytic enzymes. However, in Fig. 3(b), we show that the inulinolytic genes are not expressed in the creA strain, giving additional support that fructose does not act as an inducer for the expression of the inulinolytic enzymes. The results obtained from our transcriptional study fit well with the early observations by Vainstein & Peberdy (1991) that the invertase production in A. nidulans was the highest in sucrose medium and low in the culture fluid of fructose-grown mycelia. These findings indicate that the expression of inulinolytic enzymes is similarly regulated in A. nidulans and A. niger. As glucose, 1-kestose or 1-nystose did not induce expression of the inulinolytic genes either, we suggest a mechanism by which sucrose is transported across the plasma membrane. Once intracellular, the sucrose molecule, or a derivative of it, acts as an inducer to activate a transcription factor to drive the expression of the inulinolytic genes. Current research is aimed at the identification of transcriptional activator(s) involved in the activation of expression of inulinolytic and/ or sucrolytic enzymes, and to determine the possible role of the intracellular sucB gene during growth on inulin and sucrose.
|
ACKNOWLEDGEMENTS |
|---|
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Akimoto, H., Kiyota, N., Kushima, T., Nakamura, T. & Ohta, K. (2000). Molecular cloning and sequence analysis of an endoinulinase gene from Penicillium sp. strain TN-88. Biosci Biotechnol Biochem 64, 2328–2335.[CrossRef][Medline]
Alberto, F., Bignon, C., Sulzenbacher, G. & Henrissat, B. (2004). The three-dimensional structure of invertase (-fructosidase) from Thermotoga maritima reveals a bimodular arrangement of an evolutionary relationship between retaining and inverting glycosidases. J Biol Chem 279, 18903–18910.
Arand, M., Golubev, A. M., Neto, J. R. B. & 9 other authors (2002). Purification, characterization, gene cloning and preliminary X-ray data of the exo-inulinase from Aspergillus awamori. Biochem J 362, 131–135.[CrossRef][Medline]
Bendtsen, J. D., Jensen, L. J., Blom, N., von Heijne, G. & Brunak, S. (2004). Feature-based prediction of non-classical and leaderless protein secretion. Protein Eng Des Sel 17, 349–356.
Bennett, J. W. & Lasure, L. L. (1991). Growth media. In More Gene Manipulations in Fungi, pp. 441–447. Edited by J. W. Bennet & L. L. Lasure. San Diego: Academic Press.
Boddy, L. M., Berges, T., Barreau, C., Vainstein, C., Dobson, M. J., Balance, D. J. & Peberdy, J. F. (1993). Purification and characterization of an Aspergillus niger invertase and its DNA sequence. Curr Genet 24, 60–66.[CrossRef][Medline]
Bos, C. J., Debets, A. J., Swart, K., Huybers, A., Kobus, G. & Slakhorst, S. M. (1988). Genetic analysis and the construction of master strains for assignment of genes to six linkage groups in Aspergillus niger. Curr Genet 14, 437–443.[CrossRef][Medline]
Cairns, A. J. (2003). Fructan biosynthesis in transgenic plants. J Exp Bot 54, 549–567.
Chávez, F. P., Pons, T., Delgado, J. M. & Rodríguez, L. (1998). Cloning and sequence analysis of the gene encoding invertase (INV1) from the yeast Candida utilis. Yeast 14, 1223–1232.[CrossRef][Medline]
Chenna, R., Sugawara, H., Koike, T., Lopez, R., Gibson, T. J., Higgins, D. G. & Thompson, J. D. (2003). Multiple sequence alignment with the Clustal series of programs. Nucleic Acids Res 31, 3497–3500.
Chiou, C. Y., Wang, H. H. & Shaw, G. C. (2002). Identification and characterization of the non-PTS fru locus of Bacillus megaterium ATCC 14581. Mol Genet Genomics 268, 240–248.[CrossRef][Medline]
Coutinho, P. M. & Henrissat, B. (1999). Carbohydrate-active enzymes: an integrated database approach. In Recent Advances in Carbohydrate Bioengineering, pp. 3–12. Edited by H. J. Gilbert, G. Davies, B. Henrissat & B. Svensson. Cambridge: Royal Society of Chemistry.
de Ruiter-Jacobs, Y. M., Broekhuijsen, M., Unkles, S. E., Campbell, E. I., Kinghorn, J. R., Conteras, R., Pouwels, P. H. & van den Hondel, C. A. M. J. J. (1989). A gene transfer system based on the homologous pyrG gene and efficient expression of bacterial genes in Aspergillus oryzae. Curr Genet 16, 159–163.[CrossRef][Medline]
de Vries, R. P., van Grieken, C., van Kuyk, P. A. & Wösten, H. A. B. (2005). The value of genome sequences in the rapid identification of novel genes encoding specific plant cell wall degrading enzymes. Curr Genom 6, 157–187.[CrossRef]
Dowzer, C. E. & Kelly, J. M. (1991). Analysis of the creA gene, a regulator of carbon catabolite repression in Aspergillus nidulans. Mol Cell Biol 11, 5701–5709.
Eddy, S. R. (1998). Profile hidden Markov models. Bioinformatics 14, 755–763.
Grootwassink, J. W. D. & Hewitt, G. M. (1983). Inducible and constitutive formation of -fructofuranosidase (inulase) in batch and continuous cultures of the yeast Kluyveromyces fragilis. J Gen Microbiol 129, 31–41.
Hendry, G. A. F. & Wallace, R. K. (1993). The origin, distribution and evolutionary significance of fructans. In Science and Technology of Fructans, pp. 119–139. Edited by M. Suzuki & N. J. Chatterton. Boca Raton, FL: CRC Press.
Heyer, A. G. & Wendenburg, R. (2001). Gene cloning and functional characterization by heterologous expression of the fructosyltransferase of Aspergillus sydowi IAM 2544. Appl Environ Microbiol 67, 363–370.
Inoue, J. C., Nojima, H. & Okayama, H. (1990). High efficiency transformation of Escherichia coli with plasmids. Gene 96, 23–28.[CrossRef][Medline]
Kaplan, H. & Hutkins, R. W. (2003). Metabolism of fructooligosacchrides by Lactobacillus paracasei 1195. Appl Environ Microbiol 69, 2217–2222.
Kim, H. Y. & Choi, Y. J. (2001). Molecular characterization of cyclo-inulooligosaccharide fructanotransferase from Bacillus macerans. Appl Environ Microbiol 67, 995–1000.
Kulminskaya, A. A., Arand, M., Eneyskaya, E. V., Ivanen, D. R., Shabalin, K. A., Shishlyannikov, S. M., Saveliev, A. N., Korneeva, O. S. & Neustroev, K. N. (2003). Biochemical characterization of Aspergillus awamori exoinulinase: substrate binding characteristics and regioselectivity of hydrolysis. Biochim Biophys Acta 1650, 22–29.[Medline]
Laloux, O., Cassart, J. P., Delcour, J., Van Beeumen, J. & Vandenhaute, J. (1991). Cloning and sequence analysis of the gene encoding invertase from the yeast Kluyveromyces marxianus var. marxianus ATCC 12424. FEBS Lett 289, 64–68.[CrossRef][Medline]
L'Hocine, L., Wang, W., Jiang, B. & Xu, S. (2000). Purification and partial characterization of fructosyltransferase and invertase from Aspergillus niger AS0023. J Biotechnol 81, 73–84.[CrossRef][Medline]
Martin, I., Debarbouille, M., Ferrari, E., Klier, A. & Rapoport, G. (1987). Characterization of the levanase gene of Bacillus subtilis which sho
75 % homology; light grey, 
