| Glycobiology | Pages |
Identification and analysis of a Class 2 [alpha]-mannosidase from Aspergillus nidulans
Introduction
Results
Discussion
Materials and methods
Acknowledgements
References
Identification and analysis of a Class 2 [alpha]-mannosidase from Aspergillus nidulans
Key words: asparagine-linked/Aspergillus/alpha-mannosidase/ glycosylation
Introduction
Glycosylation, the process by which oligosaccharides are covalently linked to specific amino acids of newly synthesized proteins, can have major effects on protein structure and function. These include effects on the stability, antigenicity, and biological activity of glycoproteins (Goochee et al., 1991; Opdenakker et al., 1993); thus, protein glycosylation can be a very important factor in choosing an expression system for the production of recombinant proteins. Prokaryotic expression systems, such as Escherichia coli, produce high levels of recombinant proteins but entirely lack glycosylation and other posttranslational machinery (Kalsner et al., 1995). Eukaryotic systems are preferred for recombinant protein production because they are capable of protein glycosylation and other posttranslational modifications. Filamentous fungi of the genus Aspergillus are widely used for the expression of recombinant proteins and can produce as much as 20 g recombinant protein per liter of culture (Hintz et al., 1995). To take full advantage of the Aspergillus expression system, it would be desirable to produce glycoproteins which contain carbohydrate structures as similar as possible to the natural product. This is especially important for the production of recombinant human products of therapeutic interest such as epidermal growth factor (EGF), interleukin-6 (IL-6), and corticosteroid binding globulin (CBG) (Gwynne and Devchand, 1992). To work toward this goal, it is necessary to characterize the glycosylation pathway of secreted proteins and to understand the regulation of this process in Aspergillus.
Asparagine-linked (N-linked) protein glycosylation in higher eukaryotes is an ordered process which occurs in several stages (see reviews in Kornfeld and Kornfeld, 1985; Elbein, 1988; Herscovics and Orlean, 1993; Moreman et al., 1994). Initially, an oligosaccharide precursor consisting of three glucose, nine mannose, and two N-acetylglucosamine molecules (Glc3Man9GlcNAc2) is cotranslationally transferred to the newly synthesized polypeptide in the endoplasmic reticulum (ER). This precursor is then sequentially processed as the protein progresses through the ER and the Golgi apparatus. In the ER, [alpha]-glucosidase I and II first remove the three glucose molecules. An ER-specific Man9-[alpha]-1,2-mannosidase then removes a single mannose residue, producing Man8GlcNAc2. In the ER and Golgi, [alpha]-1,2-mannosidases remove a total of four mannose residues, yielding Man5GlcNAc2, which is the precursor for complex, hybrid, and high-mannose N-glycans. Following the addition of a single GlcNAc to Man5GlcNAc2 by GlcNAc transferase I (GnT I), mannosidase II removes two additional mannose groups, producing GlcNAcMan3GlcNAc2. Various transferases, such as GnT II, fucosyl transferase, galactosyl transferase, and sialyl transferase assemble the oligosaccharide into its final structure. In higher eukaryotes a variety of different carbohydrate units can thus be attached to a common precursor to form an array of distinct N-glycans.
It is generally accepted that the glycosylation machinery of lower eukaryotes is somewhat simpler than higher eukaryotes. Similar to higher eukaryotes, the initial precursor is processed to the Man8GlcNAc2 stage, however, in lower eukaryotes the Man8GlcNAc2 can be further mannosylated to yield N-glycans containing many mannose residues. For certain secreted and cell wall proteins, up to 200 mannose units may be added posttranslationally (Herscovics and Orlean, 1993). The precise role of the [alpha]-mannosidases in this process remains unclear. The gene for the ER-specific [alpha]-1,2-mannosidase which trims the Man9GlcNAc2 molecule to the Man8-oligosaccharide has been characterized for the yeast Saccharomyces cerevisiae (Camirand et al., 1991). This enzyme does not, however, appear absolutely necessary for further oligosaccharide mannosylation, as the Man9-oligosaccharide and the Man8-oligosaccharide are equally suitable substrates for the [alpha]-1,6-mannosyltransferase which initiates outer chain formation (Romero and Herscovics, 1989). The [alpha]-1,2-mannosidase genes have also been cloned from Aspergillus satoi and from Penicillium citrinum (two closely related fungal species; Inoue et al., 1995; Yoshida and Ichishima, 1995). Mannosidases which further process the Man8GlcNAc2 to smaller forms, such as Man5GlcNAc2, may also exist although none have yet been found. A second [alpha]-mannosidase gene, located to the vacuole, has been characterized in S.cerevisiae (Yoshihisa and Anraku, 1989, 1990). We describe the cloning and sequence characterization of a similar [alpha]-mannosidase from the filamentous fungus Aspergillus nidulans.
Results
Isolation and sequence analysis of Aspergillus [alpha]-mannosidase
The S.cerevisiae vacuolar [alpha]-mannosidase sequence (Yoshihisa and Anraku, 1989) aligned well with the rat ER/cytosolic [alpha]-mannosidase sequence (Bischoff et al., 1990) but not with other [alpha]-mannosidase sequences. Regions of shared homology were identified from this alignment of protein sequences. Conserved regions of the amino acid sequences having both a low degree of codon redundancy and a significant codon usage preference for A.nidulans were used to design semi-degenerate oligonucleotide primers for the amplification of an [alpha]-mannosidase sequence from A.nidulans strain SM222 (Table I). A codon usage table was compiled from an analysis of twenty highly expressed Aspergillus genes and the best candidate for each amino acid was selected for primer design. Amplification of A.nidulans total genomic DNA with these primers yielded a 365 bp fragment. Sequencing of this fragment revealed a single open reading frame sharing significant homology with both the rat ER/cytosolic and S.cerevisiae vacuolar protein sequences, confirming that the PCR product represented a portion of the authentic A.nidulans [alpha]-mannosidase gene. The 365 bp PCR product was then used as a probe to screen an EMBL-3 library prepared with SM222 genomic DNA. A 9685 bp BamHIfragment was recovered containing the entire coding region of the A.nidulans [alpha]-mannosidase gene as well as several thousand base pairs of flanking DNA.
Figure
The DNA sequence of the BamHI fragment was determined and a coding region 3383 bp in length was tentatively identified (Figure 1). The proposed ATG start site was embedded in the sequence GCAACCATGG, which was similar to the optimal eukaryotic initiation codon (GCC[A/G]CCAUGG), and was similar to sequences found in other filamentous fungi (ANNATGC) (Gurr et al., 1987). The DNA sequence was also determined for the 2892 bp region upstream of the putative ATG start site (Figure 1), and several motifs typical of fungal promoters were identified. The three translational reading frames of the BamHIfragment were compared to the rat and S.cerevisiae protein sequences and introns were initially identified by searching the primary sequence with 5[prime] and 3[prime] fungal intron consensus motifs (Gurr et al., 1987). The intron search sequence included a 5[prime] consensus donor site (GTNNR) and a lariat site (CTNA) positioned at least 8 bp upstream of the putative acceptor site (YAG). Three different introns were identified by the occurrence of these consensus motifs at sites where there was both a shift in the reading frame and extra residues in the expected coding region based upon the consensus sequence of the rat and yeast peptides. The positions of the introns were verified by comparing the sizes of amplification products derived from reverse transcription (RT) PCR of A.nidulans mRNA and from genomic DNA using primer pairs designed to flank the putative intron sequences. RT-PCR amplification of the mRNA using the primer pair I1F/I1R (Table I) produced a fragment which was 103 bp shorter than amplification from genomic DNA. Sequencing of the RT-PCR fragment confirmed that the portion of the gene corresponding to the first intron (Intron I, 103 bp) had been spliced out of the final transcript. Similarly, amplification with primer pair I2F/I2R and primer pair I3F/I3R (Table I) confirmed the presence of a 52 bp intron (Intron II) beginning 758 bp downstream from the start site and a 47 bp intron (Intron III) located 2785 bp downstream from the start site. This ORF produces a 1049 amino acid protein with a calculated molecular weight of 118 kDa. Six potential N-glycosylation sites (Asn-X-Ser/Thr, where X is any amino acid except proline) were identified. The putative protein sequence aligned well with the rat ER/cytosolic and S.cerevisae vacuolar sequences (Figure 2).
Figure
The Aspergillus peptide sequence was then analyzed to find regions such as ER signal sequences and transmembrane domains, which are typical of sorting or targeting signals. A Kyte-Doolittle hydropathy plot did not suggest the presence of any significant transmembrane regions. A PSORT search (Nakei and Kanehisa, 1992), which predicts protein localization based on amino acid composition, suggested that the A.nidulans [alpha]-mannosidase did not have an N-terminal signal sequence or any regions which would be typical of a vacuolar or lysosomal targeting signal.
Primer
Sequence (5[prime] [rarr] 3[prime])
AspmanF
ggggtttaagcttactgtcacatcgatacagcctggctctggcccttcnnnggac
AspmanR
ggggtttgaattcgggaggagtagccaaaggtgtcggggagccagaa
I1F
atgggtggagacgttctccgacagact
I1R
atcccagtggaattcaagccgttctct
I2F
ccctggtggtgactccattcagcc
I2R
gaacaatcaaccttgtctccgagg
I3F
gagctcgaaagcgaagtcgaatgg
I3R
gcgccttgggagctcggaggagg
KOconf-F
gacgatcccgaagatatgcgcg
KOconf-R2
cgcgccatcaggatccggtcacc
alcAprom-F
gaatagccaaaagtatccggaagccag
manR-conf
aacacttccaatcctatcacctcgc
M3SL-F
cgtaaggtgaccggatcctgatggcgcgccgtgattgaggcgacggtggtggtggcgggc
M3SL-R
aggtgtcgactggatccgtctcctgccattcgacttcgctttcg
Disruption of [alpha]-mannosidase in A.nidulans
To determine whether this [alpha]-mannosidase was essential to normal cellular function, we interrupted the coding sequence of the A.nidulans gene. A disruption cassette p8[Delta]KO (Figure 3a) containing stop codons in all three reading frames of the [alpha]-mannosidase gene was used to replace the endogenous copy of the gene. Homologous integration of the disruption cassette was confirmed by PCR amplification. The forward primer (KOconf-F) was specific for the [alpha]-mannosidase gene upstream of the recombination target while the reverse primer (KOconf-R2) was homologous to a unique region of the disruption cassette (Table I). This primer combination yielded a 1.15 kb product for transformants that integrated the disruption cassette at the authentic [alpha]-mannosidase locus. This was confirmed by sequence analysis of the 1.15 kb product. The PCR amplification product from the transformant AnKO-7 consisted of authentic [alpha]-mannosidase sequence immediately upstream of the disruption target as well as sequence derived from the disruption cassette including the stop codons and restriction sites unique to the disruption cassette. Integration of the stop codons was expected to result in the production of a truncated protein product consisting of the first 478 amino acids of the [alpha]-mannosidase protein having a calculated molecular mass of 58.6 kDa.
As illustrated by the PCR analysis, the disruption cassette integrated at the authentic [alpha]-mannosidase locus, but the cassette could also integrate at several nonhomologous sites which would not be detected by this method. To reduce potential phenotypic changes due to nonhomologous integration events (i.e., insertional mutagenesis), it was desirable to produce disruption transformants with a minimal number of nonhomologous integrations. Transformants were examined by Southern analysis to determine the number of copies of the cassette integrated at both the authentic and nonhomologous sites. Homologous replacement of the authentic DNA with the disruption cassette was expected to result in the insertion of a novel BamHI site at position +1575 of the coding region of the gene (Figure 3a). When the radiolabeled [alpha]-mannosidase DNA was hybridized to wild type (T580) and the disruptant (AnKO-7) genomic DNA cut with BamHI and BglII (Figure 4), the control T580 DNA (lane 1) showed a single 7.2 kb fragment absent from An-KO7 (lane 2). Three additional fragments of 3.0 kb, 5.9 kb, and 10.0 kb were evident in the digested DNA of the disruptant. The 3.0 kb fragment resulted from the truncation of the 7.2 kb BglII/BamHI fragment due to integration of the disruption cassette at the [alpha]-mannosidase gene, while the 5.9 kb and 10.0 kb fragments likely resulted from nonhomologous integration events. Since other transformants demonstrated numerous nonhomologous integrations (not shown), transformant AnKO-7 was chosen for further analysis.
To confirm that the gene disruption did indeed eliminate [alpha]-mannosidase activity, enzyme activities of the wild type (T580) and the disrupted strain (An-KO7) were assayed using p-nitrophenyl-[alpha]-d-mannopyranoside (PNP) as a substrate. Enzyme activity could not be detected when wild type protoplasts were osmotically lysed by the addition of water, but low levels of activity were found when protoplasts and all subcellular components were lysed by the further addition of 1% Triton X-100. Since the levels detected were still quite low, the cell lysates were concentrated on a 100 kDa filter to improve assay resolution. The [alpha]-mannosidase activity of the wild type remained constant after 38 and 48 h of growth (Figure 5). No [alpha]-mannosidase activity was detected at either time point in the disrupted strain AnKO-7. The disrupted strain did not show any alteration in growth or morphology, as compared to wild type.
Induced expression of [alpha]-mannosidase in A.nidulans
To mitigate against possible toxic effects due to increased expression of [alpha]-mannosidase in A.nidulans we used a phased expression system to overexpress the alcA-driven [alpha]-mannosidase protein. The [alpha]-mannosidase gene was fused in frame to the inducible alcA promoter (Figure 3b) and this expression cassette was introduced into A.nidulans strain T580 by cotransformation with the selectable marker pFB94. The alcA system permitted the accumulation of biomass during the first 24-36 h of growth. Upon glucose depletion alcA-driven expression was induced and expression levels monitored. Putative transformants were screened for integration of the expression cassette using an alcA promoter-specific primer, alcAprom-F, in combination with an [alpha]-mannosidase-specific return primer, manR-conf (Table I). These primers did not amplify from the endogenous [alpha]-mannosidase sequence and were diagnostic for the integration of the overexpression cassette. The induced expression transformant pAM2 was identified using this primer pair. To verify induced expression of [alpha]-mannosidase, pAM2 was cultured in inducing media and assayed for [alpha]-mannosidase activity using the aryl substrate PNP. The transformant exhibited a twofold increase in activity after 38 h of growth and a 6-fold increase in activity after 48 h, as compared to T580 (Figure 5). These transformants did not exhibit any toxic or deleterious effects, and did not show visible alterations in morphology or growth.
Sequence comparison with other [alpha]-mannosidases
The [alpha]-mannosidases have traditionally been divided into Class 1 and Class 2 sequences based on substrate specificity, inhibitor sensitivity, and sequence alignments (Daniel et al., 1994; Moreman et al., 1994). To classify the A.nidulans [alpha]-mannosidase, and to confirm previous classification schemes, a similarity matrix was generated for 18 complete [alpha]-mannosidase protein sequences, including the A.nidulans [alpha]-mannosidase, using the Matchbox alignment program (Figure 6). This program calculates the global similarity, using the BLOSUM62 scoring matrix (Henikoff and Henikoff, 1992), between each possible pair of sequences by comparing successive amino acid segments of length 9 from one sequence with all possible segments from the second sequence (Depiereux and Feytmans, 1992). The coefficient (Rij, where 0 [le] Rij [le] 1) generated for each pair of sequences represents the proportion of segments of sequence i matching with at least one segment of sequence j. This alignment method provides a more objective value for sequence similarity than other programs, since it does not depend on an alignment of all sequences to compute similarity values, but rather, scans each possible pair of sequences and computes similarity coefficients independent of other sequences used in the alignment. Several blocks of sequences exhibited high similarity to each other, but low similarity to the rest of the sequences in the matrix, as indicated by the boxed regions in Figure 6. The A.nidulans [alpha]-mannosidase, rat ER/cytosolic [alpha]-mannosidase and yeast vacuolar [alpha]-mannosidase were highly similar (Box 1), with Rij values ranging from 0.29 to 0.43. The A.nidulans [alpha]-mannosidase showed higher similarity to the S.cerevisiae vacuolar [alpha]-mannosidase (Rij = 0.43) than to the rat ER/cytosolic [alpha]-mannosidase (Rij = 0.33). This program did not find a significant global similarity between these sequences and the rest of the sequences in the matrix (Rij < 0.05). Box 2 and Box 3, containing the lysosomal [alpha]-mannosidases and mannosidase II sequences, respectively, also exhibited high internal similarity, with Rij values between 0.33 and 0.89. These two groups of sequences, however, also showed some moderate similarity to each other, as indicated by dotted lines in Figure 6 (Rij values between 0.11 and 0.16). The sequences in Boxes 2 and 3 showed very low similarity with the other sequences in the matrix (Boxes 1 and 4), with Rij values not greater than 0.05. The sequences in Boxes 1, 2, and 3 are all Class 2 [alpha]-mannosidases, as defined by Moreman et al. (1994). Box 4 consisted of all of the Class 1 [alpha]-mannosidases, and showed high internal similarity, while exhibiting a very low similarity to any of the Class 2 [alpha] -mannosidases. Certain pairs of Class 1 sequences showed extremely high similarity (i.e., mouse and human kidney [alpha]-mannosidase I: Rij = 0.93), whereas others showed more limited similarity (P.citrinum and C.elegans: Rij = 0.12), but this could be correlated to species divergence. When sequences in Box 4 were compared to other sequences in the matrix, the Rij was consistently less than 0.03.
To further resolve these sequence relationships, we compiled a dendrogram using the ClustalW algorithm (Thompson et al., 1994). Figure 7 shows the most parsimonious tree generated by the multiple alignment of all 18 [alpha]-mannosidase sequences. The Class 1 and Class 2 sequences were found to be unrelated and were represented by two separate trees. Similar to the Matchbox program, the lysosomal and mannosidase II sequences were found to be related, but tended to fall into separate subgroups (Class 2A and Class 2B). In contrast to the Matchbox program, however, the A.nidulans, rat ER/cytosolic, and yeast vacuolar [alpha]-mannosidases were found to be distantly related to the rest of the Class 2 sequences, and formed a distinct subgroup (Class 2C).
Since the Matchbox program did not find global similarity between the Class 2C and the Class 2A/2B sequences, yet the ClustalW dendrogram indicated that these sequences were distantly related, we further investigated this particular relationship between the Class 2 sequences. The Multiple Alignment Program (MAP) (Huang, 1994), which is more useful for identifying localized regions of similarity, revealed a single amino acid block which was similar between all Class 2 sequences. This block was found between amino acids 283 and 411 in the A.nidulans [alpha]-mannosidase (shown in Figure 2). This conserved 128 amino acid block confirmed that the Class 2C sequences were indeed related to the rest of the Class 2 sequences.
Figure
Discussion
The [alpha]-mannosidases play a key role in protein glycosylation, both in the modification of N-glycan chains prior to further elongation, and in the catabolism of oligosaccharides (Moreman et al., 1994). Whereas mammalian glycoproteins generally contain complex type N-glycans, which may be necessary for the correct function of the glycoprotein, recombinant production of these same glycoproteins in lower eukaryotic expression systems leads to the addition of oligomannosidic N-glycans (Kalsner et al., 1995). This can be a disadvantage of using fungal expression systems for the production of mammalian glycoproteins. We are currently characterizing the glycosylation pathways in the filamentous fungus Aspergillus nidulans so that we may better utilize this organism as an expression host. Our initial focus is on the role of the [alpha]-mannosidases in N-glycan processing.
Classification schemes have previously placed the [alpha]-mannosidases into two major groups, Class 1 and Class 2, based on biochemical properties, substrate specificity, inhibitor profiles, and sequence alignments (Daniel et al., 1994; Moreman et al., 1994). The A.nidulans [alpha]-mannosidase gene was expected to be a Class 2 [alpha]-mannosidase, as it was initially amplified based on similarity between the rat ER/cytosolic and S.cerevisiae vacuolar Class 2 [alpha]-mannosidases. To confirm this assignment, we performed a pairwise comparison of the A.nidulans [alpha]-mannosidase protein sequence to the complete protein sequences of 17 Class 1 and Class 2 [alpha]-mannosidases using the Matchbox algorithm (Depiereux and Feytmans, 1992). An advantage of this algorithm is that, unlike many sequence alignment programs, it does not rely on a user-defined gap penalty. Sequence similarity was computed independent of the multiple sequence alignment thus producing the same result for any pair of sequences independent of the number of sequences included in the analysis. We found that the Class 1 sequences, while being highly similar to each other, were not related to the Class 2 [alpha]-mannosidases. The lysosomal [alpha]-mannosidases and mannosidase II sequences were found to be similar to each other, and tended to form two separate subgroups of higher similarity. Interestingly, the rat ER/cytosolic, S.cerevisiae, and A.nidulans [alpha]-mannosidases, while highly similar to each other, were not found to be globally similar to any of the other sequences. This analysis, while providing excellent overall sequence similarity coefficients, was not designed to locate localized regions of similarity, which may have been present in more distantly related sequences.
Figure
To identify localized regions of similarity, we utilized two alignment programs, ClustalW (Thompson et al., 1994) and the Multiple Alignment Program (MAP) (Huang, 1994). Both of these programs detected small, yet significant regions of similarity between all of the Class 2 sequences, supporting previous alignment based studies which suggested that these enzymes were all related (Daniel et al., 1994; Moreman et al., 1994). More recently, Henrissat and Bairoch (1993, 1996) also found that the Class 2 enzymes tended to belong to a single group of glycosylhydrolases (#38), whereas the Class 1 [alpha]-mannosidases belonged to a separate group (#47). The authors identified a protein domain which was common to all the Class 2 enzymes and likely represented a common catalytic domain. Our analysis, which included the A.nidulans [alpha]-mannosidase, emphasized the three subgroups of Class 2 [alpha]-mannosidases (2A, 2B, and 2C) which exhibited distinct sequence and biochemical properties. A dendrogram, compiled using the ClustalW alignment, showed that the Class 2C sequences were quite distantly related to the rest of the Class 2 [alpha]-mannosidases (Figure 6). Many conserved regions existed within the Class 2C enzymes (Figure 2) which were not found in the Class 2A or 2B enzymes. While the Matchbox algorithm was useful for objectively identifying overall sequence similarities and for defining the various groups of enzymes, more refined sequence alignment programs were required to detect the similarities between the more distantly related groups.
Figure . Matchbox sequence similarity matrix. Coefficients of similarity (Rij) of [alpha]-mannosidase protein sequences using Matchbox algorithm, where Rij is the proportion of amino acid segments (length 9) of sequence i matching with at least one segment of sequence j. Boxed regions represent groups of sequences which show a relatively high degree of peptide sequence similarity to each other, but not to other sequences in the matrix. Box 2 and Box 3 also show some limited similarity (dotted lines), and are more similar to each other than they are to Box 1 or Box 4. Protein sequences are numbered as follows: 1, A.nidulans [alpha]-mannosidase; 2, rat ER/cytosolic [alpha]-mannosidase (Bischoff et al., 1990); 3, S.cerevisiae vacuolar [alpha]-mannosidase (Yoshihisa and Anraku, 1989); 4, Dictyostelium discoideum lysosomal [alpha]-mannosidase (Schatzle et al., 1992); 5, human lysosomal [alpha]-mannosidase (Nebes and Schmidt, 1994); 6, mouse lysosomal [alpha]-mannosidase (Merkle and Moreman, 1993); 7, Drosophila melanogaster Golgi mannosidase II (Foster et al., 1995); 8, human Golgi mannosidase II (Misago et al. 1995); 9, mouse Golgi mannosidase II (Moreman and Robbins, 1991); 10, mouse Golgi mannosidase IA (Lal et al., 1994); 11, rabbit liver mannosidase I (Lal et al., 1994); 12, human Man9-mannosidase (Bause et al., 1993); 13, mouse Golgi mannosidase IB (Herscovics et al., 1994); 14, D.melanogaster mannosidase I (Kerscher et al., 1995); 15, S.cerevisiae Man9-mannosidase (Camirand et al. 1991); 16, P.citrinum mannosidase I (Yoshida and Ichishima, 1995); 17, A.satoi mannosidase I (Inoue et al. 1995); 18, Caenhaboriditus elegans mannosidase I (Wilson et al., 1994).
Figure
Although the Class 2C [alpha]-mannosidases showed a high degree of sequence similarity, the cellular locations and functions of the enzymes were variable. Although we were unable to conclusively determine the cellular location of the A.nidulans [alpha]-mannosidase, the protein was likely located in a subcellular organelle, since lysis of the cell and all subcellular components with Triton X-100 was necessary to release enzyme activity. The A.nidulans protein lacked a typical signal sequence or major transmembrane domain raising questions as to its method of localization. The rat and S.cerevisiae genes also did not contain typical signal sequences (Bischoff et al., 1990; Yoshihisa and Anraku, 1990), though the S.cerevisiae protein was transported independently of the standard ER/Golgi sorting mechanism. As was reported for the S.cerevisiae vacuolar [alpha]-mannosidase (Yoshihisa and Anraku, 1989, 1990), the A.nidulans [alpha]-mannosidase was not essential for normal cellular function, as disruption of the gene did not have any visible effect on growth or morphology. The functions of the S.cerevisiae and A.nidulans [alpha]-mannosidases, however, were not determined. The rat ER/cytosolic enzyme, to which the A.nidulans gene showed a high degree of sequence similarity, is involved in oligosaccharide catabolism in both the ER and cytosol (Tulsiani and Touster, 1987; Bischoff et al., 1990; Haeuw et al., 1991; Moore and Spiro, 1994; Grard et al., 1996), and it is possible that the A.nidulans [alpha]-mannosidase plays a similar role. In order to more fully understand the role of the A.nidulans [alpha]-mannosidase in protein glycosylation, more detailed analysis of the N-glycans produced by these transformants is required. It is possible that there are yet other Class 2 [alpha]-mannosidases in A.nidulans. Since lower eukaryotes do not produce complex N-glycans, and mannosidase II is only necessary for complex N-glycan synthesis, it seems likely that this enzyme would not be found in A.nidulans. Since lysosomal [alpha]-mannosidase activity is necessary for oligosaccharide catabolism in higher eukaryotes, it would be reasonable to expect to find a similar activity in A.nidulans. The substrate p-nitrophenyl-[alpha]-d-mannopyranoside (PNP) has been used to assay Class 2 [alpha]-mannosidase activity. Surprisingly, when the A.nidulans Class 2C [alpha]-mannosidase was disrupted the PNP assay activity was reduced to zero. Based on examples of molecular sizes known for mannosidases in other species, the concentrated lysate was expected to also contain a lysosomal [alpha]-mannosidase (MW = 115-121 kDa). It is possible that the activity of the lysosomal [alpha]-mannosidase was below the detection levels of our particular assay, but, alternatively, it is possible that A.nidulans contained a single Class 2 [alpha]-mannosidase. We are currently attempting to isolate a lysosomal [alpha]-mannosidase from A.nidulans by designing degenerate PCR primers from the known lysosomal [alpha]-mannosidase gene sequences. Protein glycosylation is an important posttranslational modification which can affect numerous properties of the protein and may be very important when using heterologous expression systems for the recombinant expression of proteins. Although the glycosylation pathways of higher eukaryotes are well characterized, other systems remain relatively undefined. We have cloned a Class 2 [alpha]-mannosidase from A.nidulans which appears to be involved in N-glycan catabolism. Induced expression of [alpha]-mannosidase resulted in the accumulation of 2- to 6-fold higher levels of the enzyme which did not affect the growth rate or morphology of the fungus, hence this system can provide a recombinant source of the A.nidulans [alpha]-mannosidase for purification. Biochemical characterization and cellular localization of this enzyme in Aspergillus will be necessary to determine its exact role in protein glycosylation. We are attempting to identify other genes in the glycosylation pathway in this organism, particularly [alpha]-mannosidases and mannosyltransferases, to further characterize these pathways in lower eukaryotes.
Materials and methods
Strains, media, and growth conditions
The full-length [alpha]-mannosidase gene was isolated from Aspergillus nidulans spore color mutant SM222. The A.nidulans expression host T580, a derivative of strain FGSC4 (Fungal Genetics Stock Center) is a uridine auxotroph (ura-) which was used for expression studies.Cultures were grown in CYM liquid media (10 g glucose; 2 g bactopeptone; 1.5 g casamino acids; 1 g yeast extract; 10 ml 100× salt solution; 1 ml 1000× trace elements; 10 ml 100× vitamin solution; and 10 ml 100× adenine solution, per liter). Stock solutions (100× salt, 100× vitamin, 1000× trace elements, 100× adenine) were described in Kalsner et al. (1995). Nontransformed T580 strains were supplemented with 10 mM uridine. Protoplasts which integrated the pyrG selectable marker (pFB94) were selected for uridine prototrophy on minimal media (1 g fructose; 12 g threonine; 10 ml 100× salt solution; 1 ml 1000× trace elements; and 0.6 g NaNO3, per liter) supplemented with 0.6 M sucrose as an osmoticum. Selected transformants for overexpression of the mannosidase were grown in liquid yeast-fructose-threonine (YFT) (5 g yeast extract; 2 g fructose; 12 g threonine; 10 ml 100× salt solution; 1 ml 1000× trace elements; and 0.6 g NaNO3, per liter). Strains were maintained on CYM agar, and spore suspensions were obtained by washing cultured CYM agar plates with 8 ml 0.001% Tween 80. Mycelium for DNA isolations was obtained by inoculating 500 ml liquid CYM with 108 spores, and incubating 24 h at 30°C with constant agitation (200 r.p.m.). Expression cultures were grown by inoculating 50 ml liquid YFT media with 108 spores, and incubating at 30°C with constant agitation (200 r.p.m.) for the duration of growth.
Oligonucleotide primer design
The forward primer (Table I) was designed by reverse translation of the protein sequence CHIDTAWLWPFXET, which was perfectly conserved between the rat and yeast sequences except at position X. The primer was made fully redundant at the three positions corresponding to amino acid position X. This forward primer contained a short 5[prime] tail with an embedded HindIIIsite preceding the coding sequence to enable directional cloning of the PCR product. The return primer was the complement of the reverse translation of the amino acid sequence FWLPDTFGYSS, which was perfectly conserved between the rat and yeast sequence except for the last serine (Figure 2). This primer was also tailed at the 5[prime] end with a short sequence containing an embedded EcoRI site to facilitate the cloning.
DNA isolation and PCR amplification of [alpha]-mannosidase
Total genomic DNA was extracted from finely ground freeze-dried mycelium. Approximately 400 mg of mycelium was vortexed with 2.5 ml of 50 mM EDTA, 0.2% SDS; centrifuged for 10 min; and then 85 µl of 3 M KOAc; 5 M acetic acid added to the supernatant. Following a 20 min incubation on ice, the suspension was recentrifuged and DNA was isopropanol precipitated from the supernatant. After resuspension in 100 µl TE (10 mM Tris pH 7.5; 1 mM EDTA), the DNA was extracted once with phenol, twice with chloroform/isoamyl alcohol (24:1) and ethanol precipitated. Each PCR reaction consisted of 10-100 ng of genomic DNA, 50 pmol of each primer, 10 mM Tris-HCl (pH 8.3), 1.5 mM MgCl2, 50 mM KCl, 0.01% gelatin, 0.1% Triton X-100, 200 µM each of dATP, dCTP, dTTP, and dGTP and 2 U Taq DNA polymerase (Perkin-Elmer) in a final volume of 100 µl. Amplification was performed in two stages using a RoboCycler (Stratagene) thermal cycler. Five cycles at a lower stringency (94°C for 60 s, 55°C for 120 s, 72°C for 180 s) were followed by 25 cycles at higher stringency (94°C for 60 s, 60°C for 120 s, 72°C for 180 s). The PCR product from the tailed amplification of the A.nidulans (SM222) nuclear DNA was directionally cloned (EcoRI/HindIII) into pTZ18R prior to sequencing. The derived amino acid sequences of all three reading frames of the PCR-amplified DNA were compared to published [alpha]-mannosidase protein sequences to confirm that the PCR fragment represented a portion of the authentic [alpha]-mannosidase gene.
Library construction and screening
The [alpha]-mannosidase-specific PCR fragment was digoxigenin-labeled (DIG) by reduction of the dTTP concentration to 190 µ[Lambda] and the inclusion of 10 µM DIG-11-dUTP (Boehringer Mannheim) in a second PCR reaction using the cloned DNA as a target. The DIG-labeled DNA probe was used to identify the full length [alpha]-mannosidase gene by Southern hybridization. A partial library of A.nidulans (SM222) sequences was constructed by digesting genomic DNA with BamHI and ligating the resulting fragments into the similarly digested lambda DNA vector EMBL-3. Concatamers of the ligated DNA were packaged using the Gigapack II (Stratagene) in vitro packaging system. Approximately 105 recombinant lambda plaques were immobilized on nylon membranes (Genescreen Plus, Dupont) and hybridized with the DIG-labeled PCR fragment. A single strongly hybridizing lambda clone was subcloned into the BamHI site of pTZ18R (p[alpha]-mann18) in preparation for sequencing.
Sequence analysis
The DNA sequence was determined by the dideoxynucleotide method using the T7 sequencing kit (Pharmacia). Various restriction fragments of p[alpha]-mann18 were subcloned and individually sequenced using M13 universal primers. Overlapping regions were sequenced with synthetic specific primers and in all cases, both DNA strands were sequenced. Generunner sequence analysis software (Hastings Software Inc., Hastings-on-Hudson, NY) was used to identify open reading frames, generate alignments, and deduce polypeptides from the primary DNA sequence. FASTA searches of the GenPept and SwissProt databases (GenBank, Los Alamos) were made using the deduced polypeptides. The PSORT (Nakei and Kanehisa, 1992) program was used for prediction of protein localization sites in the deduced polypeptide. DNAStar sequence analysis software (DNAStar, Madison, WI) was used to compute amino acid similarities and to generate putative phylogenetic relationships. Intron positions were confirmed by reverse transcription (RT)-PCR of [alpha]-mannosidase cDNA. Total RNA was isolated from A.nidulans mycelium using a single-step guanidine isothiocyanate extraction (TRIzol reagent, Life Technologies). The RNA was treated with DNAase, extracted twice with chloroform containing 4% isoamyl alcohol, ethanol precipitated and reverse-transcribed to cDNA using oligo dT primers (Stratascript RT-PCR kit, Stratagene). This cDNA was subsequently amplified using primer pairs which flankedthe putative splice sites and compared to fragments obtained by amplifying genomic DNA with the same primer pairs. Sequencing of the cDNA fragments was used to confirm the exact location and size of the predicted introns.
Construction of disruption cassettes
The disruption cassette p8[Delta]KO (Figure 3a) was derived from a 2.2 kb EcoRI-SalI [alpha]-mannosidase fragment. This 2.2 kb fragment was modified by the addition of diagnostic restriction enzyme sites, as well as stop codons (TGA) in each of the three reading frames. Unique sites were added by tailed PCR amplification. The 5[prime] overhang of the forward primer M3SL-F (Table I) encoded unique BstEII, BamHI, and AscI restriction sites, and the three stop codons, while the remainder of the primer encoded authentic sequence which primed immediately downstream of an authentic BstEII site. The reverse primer M3SL-R contained an embedded SalI restriction site, and annealed just upstream of the authentic SalI restriction site. The 1.1 kb PCR reaction product was used to directly replace the 1.1 kb BstEII-SalI fragment of p8SR, using the BstEII and SalI sites for directional ligation of the fragment into similarly digested p8SR vector. Restriction analysis of the plasmid and sequencing of the inserted DNA region confirmed that the restriction sites and stop codons were intact.
Construction and induction of inducible expression cassette
The inducible expression cassette (palcAman) was created by direct fusion of the full-length [alpha]-mannosidase coding region with the inducible alcA promoter (Figure 3b) via NcoI insertion of the [alpha]-mannosidase gene into the expression cassette. Sequence analysis verified proper fusion of the promoter and coding region for expression of the gene. The alcA promoter is subject to carbon catabolite repression and transcription initiation is enhanced via the transcriptional activator AlcR in the presence of ethanol. In the presence of glucose, transcription from alcA-driven genes was repressed by interaction of the negative regulatory protein CreA with the alcA promoter (Hintz and Lagosky, 1993). The alcA-driven genes were regulated by controlling glucose levels during fermentation. Transformants were grown first in the presence of 1% glucose to repress the alcA promoter and prevent any possible toxic effects which may have arisen from overexpression of the gene. Selected transformants which had integrated the inducible expression cassette were then grown in glucose limiting conditions in the presence of an inducer, allowing for overexpression of the cytosolic [alpha]-mannosidase.
Protoplasting and transformation
Protoplasts were prepared according to the method of Fincham (1989), using Sigma Lysing Enzyme for cell wall digestion. Protoplasts of strain T580 (ura-) were cotransformed with 1 µg of the selectable marker pFB94, which converts transformed cells to uridine prototrophy, and with 1 µg of the disruption or overexpression (p8[Delta]KO or palcAman). Transformants were initially screened for integration of pFB94 by selection on minimal media and were then assayed for cotransformation by PCR analysis. Genomic DNA of putative transformants was prepared by the method of Cenis (1992), and screened using primer pairs diagnostic for the nonselectable cotransforming plasmids. Gene disruption was confirmed by Southern hybridization (Genescreen, Dupont).
Mannosidase assays
Mycelium was grown in liquid YFT cultures, harvested on Miracloth (CalBiochem) by suction filtration, and washed once with maleate buffer (0.4 M MgCl2, 0.05 M maleate, pH 5.8). Cell walls were digested with 10 mg/ml Novozyme 234 (Interspex) in maleate buffer with constant shaking (200 r.p.m.) at 30°C. Protoplasts were purified by prescreening through a 100 µm mesh followed by centrifuging at 3000 r.p.m. for 15 min at 4°C, washed twice with 10 ml ice-cold STC (1.2 M sorbitol, 100 mM Tris-HCl pH7.4, 10 mM CaCl2), pelleted, and resuspended a final volume of 1 ml STC. The protoplasts were lysed on ice by the addition of an equal volume of 10 mM Tris, pH 7.6; 1% Triton X-100, and the protoplast debris was pelleted by centrifugation (3000 r.p.m., 4°C). The supernatant proteins were size selected with an Amicon stirred ultrafiltration cell model 8050, fitted with a 100 kDa membrane. Samples of proteins >100 kDa were concentrated to a final volume of 1.5 ml in 0.01 M sodium acetate pH 4.5. Mannosidase activity was quantified as described by Matta and Bahl (1972). Equal volumes of concentrate and 5 mM PNP (p-nitrophenyl-[alpha]-d-mannopyranoside) in 0.01 M sodium acetate, pH 4.0 were incubated at 37°C for 60 min. The reaction was terminated with 1 ml of 0.2 M sodium carbonate, pH 11.8 and the absorbance of the yellow chromogen measured at 420 nm. One unit of enzyme activity (U) was defined as that which releases 1 µmol of p-nitrophenol per hour.
Sequence alignments
Matchbox sequence alignments were performed with 17 [alpha]-mannosidase sequences obtained from the GenBank database. Sequence similarities were identified using the web-based Matchbox program (http://dot.imgen.bcm.tmc.edu:9331/multi-align/multi-align.html), utilizing the BLOSUM62 similarity matrix, and default parameters. Sequence alignments using the ClustalW algorithm were performed with DNAStar sequence analysis software (DNAStar, Madison, WI), utilizing the PAM250 similarity matrix and default alignment parameters. This program was used to generate the most parsimonious phylogenetic tree for the sequences. The Multiple Alignment Program (MAP) was used to find localized regions of similarity, using the web based server (http://dot.imgen.bcm.tmc.edu:9331/multi-align/multi-align.html) with default parameters and the BLOSUM62 similarity matrix.
Acknowledgements
This research was supported by the Natural Sciences and Engineering Research Council of Canada (0GP0138387).
References
1To whom correspondence should be addressed
This page is run by Oxford University Press, Great Clarendon Street, Oxford OX2 6DP, as part of the OUP Journals
Comments and feedback: www-admin{at}oup.co.uk
Last modification: 18 Jan 1998
Copyright© Oxford University Press, 1998.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  Y. Li, W. Fang, L. Zhang, H. Ouyang, H. Zhou, Y. Luo, and C. Jin Class IIC {alpha}-mannosidase AfAms1 is required for morphogenesis and cellular function in Aspergillus fumigatus Glycobiology, June 1, 2009; 19(6): 624 - 632. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. Li, L. Zhang, D. Wang, H. Zhou, H. Ouyang, J. Ming, and C. Jin Deletion of the msdS/AfmsdC gene induces abnormal polarity and septation in Aspergillus fumigatus Microbiology, July 1, 2008; 154(7): 1960 - 1972. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Angelov, M. Putyrski, and W. Liebl Molecular and Biochemical Characterization of {alpha}-Glucosidase and {alpha}-Mannosidase and Their Clustered Genes from the Thermoacidophilic Archaeon Picrophilus torridus. J. Bacteriol., October 1, 2006; 188(20): 7123 - 7131. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. Reggiori and D. J. Klionsky Autophagy in the Eukaryotic Cell Eukaryot. Cell, February 1, 2002; 1(1): 11 - 21. [Full Text] [PDF]
|
|
|
|
 |
|
 M. U. Hutchins and D. J. Klionsky Vacuolar Localization of Oligomeric alpha -Mannosidase Requires the Cytoplasm to Vacuole Targeting and Autophagy Pathway Components in Saccharomyces cerevisiae J. Biol. Chem., June 1, 2001; 276(23): 20491 - 20498. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||