Introduction of the Aspergillus fumigatus α-1,2-mannosidase MsdS into Trichoderma reesei improves the ligno-cellulose degradation


 Background: a-1,2-Mannosidase is an important enzyme essential for N-glycan processing and plays a significant role in the biosynthesis and organization of fungal cell wall. Lacking of α-1,2-mannosidase leads to cell wall defect in yeast and filamentous fungi. In Trichoderma reesei, a fungus known to be non-toxic to human, its N-glycan on secreted glycoprotein is Man8GlcNAc2, which is different from that in Aspergillus fumigatus. To evaluate the significance of the N-glycan processing in T. reesei, in this study A. fumigatus α-1, 2-mannosidase MsdS, an enzyme that cleaves N-linked Man8GlcNAc2 in Golgi to produce Man6GlcNAc2 on secreted glycoprotein, was introduced into T. reesei.Results: The msdS-expressing strain Tr-MsdS produced a major glycoform of Man6GlcNAc2 on its secreted glycoproteins, instead of Man8GlcNAc2 in the parent strain. Although the cell wall content of msdS-expressing strain Tr-MsdS was changed, it appeared that the cell wall integrity was not affected. However, phenotypes such as increased conidiation, multiple budding and random branching were observed in strain Tr-MsdS. In addition, expression of MsdS into T. ressei also affected protein secretion and improved the ligno-cellulose degradation of T. reesei.Conclusions: Our results indicate that processing of the N-glycan is species-specific and plays an important role in protein secretion in T. reesei, specially cellulases. Also, our results provide a new strategy to improve cellulases production by interfering the N-glycan processing in T. reesei.

that O-glycosylation affects not only the function of these enzymes, such as proteolytic resistance, thermostability and cellulose binding [7], but also their expression and secretion [8][9]. Interestingly, either increase or decrease of O-glycosylation level leads to an increase of secreted proteins [10][11].
However, it keeps unknown how N-glycan processing affects protein secretion.
In eukaryotic cells, secreted proteins are synthesized in the endoplasmic reticulum (ER) and transported to Golgi apparatus. In the course of the secretory pathway, N-glycosylation is initiated in the lumen of the ER by transferring of Glc 3 Man 9 GlcNAc 2 from dolichol pyrophosphate to the nascent polypeptides. Once the Glc 3 Man 9 GlcNAc 2 is transferred to proteins, N-Glycan processing is initiated sequentially in the ER and Golgi by two ER α-glucosidases and various 1,2-α-mannosidases [12][13]. In mammalian cells, Man 9 GlcNAc 2 is converted to Man 5 GlcNAc 2 by the action of ER and Golgi amannosidases, and Man 5 GlcNAc 2 is the precursor for complex, hybrid, and high-mannose type of Nglycans [12]. Golgi a-mannosidase is a Class-I α-mannosidase responsible for cleavage of α-1,2-linked D-mannoses [14][15][16]. In vitro analysis reveals that this enzyme converts Man 8 GlcNAc 2 or a mixture of Man [6][7][8][9] GlcNAc 2 oligosaccharides to the respective Man 5 GlcNAc 2 structures [15,17].
Previously we have shown that in filamentous fungi Aspergillus fumigatus a-mannosidase MsdS is responsible for processing of N-linked Man 8 -9 GlcNAc 2 to Man 6 GlcNAc 2 , which is the glycoform on secreted glycoproteins [17]. Deletion of the a-mannosidase gene msdS results in a conversion of Nglycan attached to secreted glycoproteins from Man 6 GlcNAc 2 to Man 8 GlcNAc 2 and causes defective cell wall and abnormal polarity [17]. On the other hand, it is interesting to note that the N-glycan on secreted glycoproteins of T. reesei is Man 8 GlcNAc 2 [18], which is the same as that of the A. fumigatus msdS-knockout mutant, suggesting that N-glycan processing is different in these two species.
In attempt to evaluate the effect of the altered N-glycan processing on T. reesei, in this study a msdSexpressing strain Tr-MsdS was constructed by introducing of the A. fumigatus msdS gene into T.
reesei. As expected, the main N-glycan glycoform was converted from Man 8  coli [17], however, the specificity of the antibody was not good (data not shown). Therefore, we further analyzed proteins extracted from Tr-MsdS strain with LC-MS/MS. To this end, the secreted proteins of Tr-MsdS strain were separated by polyacrylamide gels and protein bands corresponding to 40-60 kDa were subjected to LC-MS/MS analysis. As a result, 16 peptides were identified as fragments of MsdS (gi159131578) ( Fig.S1 and Fig.S2). Although a MsdS homolog (TRIREDRAFT_45717) was found in T. reesei (Fig.S3), no peptide was identified. These results clearly demonstrated that MsdS was successfully expressed in Tr-MsdS strain, while the T. reesei MsdS homolog was not expressed.

Phenotype of Tr-MsdS strain.
The growth kinetics of T. reesei strains were studied on solid media by measuring colony diameter.
The result showed that 28-32°C were the favorable temperatures for both parent and Tr-MsdS strains, however Tr-MsdS strain was observed to be slightly temperature-sensitive at 42°C as compared with its parent strain (Fig.S4). Additionally, at 32 °C, the conidia produced by parent strain was more than that produced by Tr-MsdS strain (47%, 61% and 38% at 12 h, 24 h and 48 h, respectively), whereas after 72 h the conidia produced by Tr-MsdS strain appeared higher. This suggests that expression of MsdS leads to a slightly slower growth rate at early phase ( Table 1).
As shown in Fig.3A, in the presence of antifungal reagent, hyphal growth of Tr-MsdS strain was not affected at 28°C, 32°C or 37°C. When observed under transmission electron microscope (TEM) (Fig.3B), Tr-MsdS strain grown at 32°C showed 30% more thickened hyphal cell wall as compared with its parent strain. When the temperature was raised to 37°C, the thickness of the Tr-MsdS mycelia cell wall was only 13 % more as compared with its parent strain. In addition, the conidial cell wall of Tr-MsdS strain formed at 32°C was 9% thicker than at 37°C, whereas the cell wall of parent strain showed 27% less thick at 37°C as compared with 32°C. Interestingly, Tr-MsdS strain showed less dense and irregularly scattered filamentous material around spore and hypha, whereas its parent strain showed denser, regular and prominent filamentous material surrounding spore and hyphae.
These results indicated that hyphal cell wall thickness was changed in Tr-MsdS strain and altered at 37°C.
The cell wall components including glycoprotein, glucan and chitin were further analyzed ( Table 2).
The glycoprotein content was reduced by 30-40% in Tr-MsdS strain at both 32°C and 37°C. The cell wall chitin in Tr-MsdS strain was reduced by 15% at 32°C and by 10% at 37°C. Interestingly, α/βglucan was found to be increased by 25% in Tr-MsdS strain. These phenomena suggested that the expression of MsdS resulted in an increase of α/β-glucan in T. reesei. Previously, 10-27% reduction of α-glucan, mannoprotein, β-glucan and chitin were observed in the A. fumigatus DmsdS mutant, while 25%, 33% and 55% in α-glucan, β-glucan and chitin respectively was observed when the temperature was increased [17]. On the other hand, the cell wall glycoprotein was reduced in Tr-MsdS strain ( Fig.3C and Table S1). Previous investigation in A. fumigatus showed that unfolded protein response (UPR) is activated by reduced N-glycosylation and overexpressed cell wall protein and chitin [17]. In this study, UPR transcription activator factor hac1, was found to be reduced (Fig.4). This could provide some information that UPR is not induced in Tr-MsdS strain and there is no protein misfolding, which could have improved the growth and secretion in Tr-MsdS strain followed by enhanced α/β-glucan content. In addition, the enhanced expression of sec61 and rho3 which function to mediate protein translocation across ER and control the cell shape, respectively, has been clearly observed in Tr-MsdS strain. This provides support to the result that there is a change of the genes involved in secretory pathway of T. reesei and hence cell wall component has been altered when the msdS gene was introduced.
When the dormant conidia start to germinate in A. fumigatus, a series of nuclear division occur which is accompanied with an order of morphological events (switch from isotropic to polar growth) including appearance of first/second germ tubes and septation [19][20]. In the A. fumigatus DmsdS mutant random budding and septation at an early stage of germination were observed [17]. In A.
nidulans, the temperature-sensitive mutant which is unable to switch from isotropic to polar growth was also reported although multiple points of polarity were established [19][20]. In some other studies, the abnormal morphology with balloon-like structures, swollen hyphae tips and altered cell cycle was described with a chitin synthase deficient mutant of F. oxysporum as well as T. reesei [21].
In the present study, the parent and Tr-MsdS strain were grown in media containing glucose as the carbon source and observed for the germination of conidium. Fig.5 and Table 3 show that both parent and Tr-MsdS strain produce first germ tube at 6 h and initiation of the second germ tube at 7 h. In parent strain the germination showed a typical and regular branched apical growth usually at the early stage of germination, while less branched hyphae at the later stage of germination. However, in Tr-MsdS strain slightly swollen hyphal tips were observed together with multiple budding sites and random branching. In addition, during early stage, the first germ tube grew longer and then lateral germ tube appeared in Tr-MsdS strain. The spores of parent strain germinated in a typical polarized pattern in which the first and second germ tube occurred at 5 h and 6 h, respectively, and the first septation occurred after 9 h. Surprisingly, the occurrence of the first and second germ tube was rapid until 8 h in parent strain and then slowed down, whereas Tr-MsdS strain started to generate the second germ tube more rapidly after 8 h. The chitin accumulation was observed at the basal and lateral portion of hyphae. Very few septation occurred in both parent and Tr-MsdS strain. Previous studies have also reported that there is a proportional correlation of increased level of chitin with the level of glucan depicting that the cells will turned on the cell wall compensatory mechanism as a reaction under stress [21]. These results clearly demonstrated that insertion of the msdS gene led to more branched and random budding at later stage of germination.

Glycosylation and protein expression in Tr-MsdS strain.
To compare the N-glycosylation between parent and mutant strain, the N-glycans were released from membrane and secreted proteins with PNGase F and analyzed by MALDI-TOF. As shown in Fig.6, signals of Man 5-8 GlcNAc 2 were detected in parent strain, among them Man 8 GlcNAc 2 was the major glycoform. Although signals corresponding to Man 5-9 GlcNAc 2 were detected in Tr-MsdS strain, the glycoform Man 8 GlcNAc 2 was greatly reduced, suggesting an action of MsdS on N-glycans.
As shown in Fig.2B, the intracellular, cell wall, membrane and secreted proteins of both strains showed no difference. However, the MS/MS analysis of secreted protein identified some important proteins involved in secretory and regulatory pathway of T. reesei, such as 14-3-3-like protein (gi|12054274), HSP70 (gi|30961863) and small GTPase of the Rab/Ypt family (gi|340519278) ( Table   S2). Saloheimo and Pakula have clearly explained about the secretory pathway in filamentous fungi including T. reesei [22]. Based on the T. ressei protein secretory pathway (Fig.7), the results were further analyzed by transcriptional expression (Fig.4). Interestingly, it was observed that three genes encoding Sec61 (major component of protein translocation complex in ER membrane), Ftt1 (a 14-3-3 type involved in last step of secretory pathway) and Rho3 (Ras-type GTPase involved in cell polarity and vesical fusion with plasma membrane) were found to be expressed higher in Tr-MsdS strain as compared with its parent strain. Whereas expressions of the genes encoding Hac1, Rab5, Snc1 and Ypt1 were reduced in Tr-MsdS strain. These results suggested that expression of MsdS affected the protein transportation and secretion in T. reesei.

Cellulase and β-mannanase in Tr-MsdS strain.
In the present study, cellulase and β-mannanase activity were analyzed using carboxymethyl cellulose and locust bean gum as a substrate, respectively. Both activity and hydrolysis yield were found to be higher in Tr-MsdS strain as compare with its parent strain. As shown in Fig.8, glucose yield was 17.5%, 12.6% and 9.9% higher in Tr-MsdS strain at 30°C, 40°C and 50°C, respectively. Also, mannose yields in Tr-MsdS strain were 27.1%, 25.7% and 32.2% higher than that in the parent strain at 30°C, 40°C and 50°C, respectively. These observations suggest that introduction of MsdS into T.
ressei not only affects the glycosylation and cell wall synthesis, but also improves the ligno-cellulose degradation. Further, the yield was found to be increased at temperature up to 50°C (industrially applicable temperature). This could also somehow attribute to be applied in industrial purpose to degrade complex substrate after optimizing proper condition for efficient degradation. The Tr-MsdS in the present study somehow acted as a fungus constructed with additional mannan degrading enzyme providing improved hydrolysis of locust bean gum or cellulose to some extent.

Discussion
The primary steps of protein N-glycosylation is common among fungi and mammals including the sitespecific transfer of Glc 3 Man 9 GlcNAc 2 from ER lumen to the de novo synthesized protein followed by subsequent trimming by glucosidases I and II and a specific ER-residing a-1,2-mannosidase to form Man 8 GlcNAc 2 structures (isomer Man8B), which later export the predominant Man8-isomer to the Golgi [23]. In human Golgi α-1,2-mannosidase (IA-IC) removes Man to yield the Man 5 GlcNAc 2 structure (the precursor for complex N-glycans) [24][25][26], whereas in S. cerevisiae [24], the N-glycan processing, involves the addition of numerous mannose sugars throughout the entire Golgi, often leading to hyper-mannosylated N-glycan structures. Previous studies have explained the consequences of Golgi α-mannosidase I activity vary in different species. The deficiency of αmannosidase results in the lethal disease called mannosidosis in human [27] and cattle [28], and outer chain synthesis in S. cerevisiae [29].
In T. reesei, the most common N-glycan structure is found to be mono-glucosylated high-mannose glycan GlcMan 7-8 GlcNAc 2 , which is similar to a transient structure of GlcMan 9 GlcNAc 2 found in the ER [30]. However, a number of other N-glycan structures also have been reported in different strains. For example, the N-glycan of Cel7A in T. reesei strain QM9414 and ALKO2877 is single GlcNAc residues in three (Asn45, Asn270, and Asn384) out of four potential sites characterized in the catalytic domain [31], while Garcia et al. also identified heterogeneous N-glycans on endoglucanase I (Cel7B) from strain QM9414 [18]. In hyper-producing strain Rut-C30, more complex N-glycan structures, including single GlcNAc and Hex 8-9 GlcNAc 2 structures, are found on cellulasese Cel7B, Cel6A and Cel5A [32][33].
These observations suggest that the N-glycan structures on cellulases produced by different mutant strains are different. Although T. reesei α-1,2-mannosidase can sequentially cleave all α-1,2-linked mannose sugars from a Man 9 GlcNAc 2 oligosaccharide giving broad substrate specificity that sterically allowed different oligosaccharide conformations [14], it is also possible that some enzymes in glycosylation pathway or some glycosylation sties on cellulases were mutated as these mutants were screened for cellulases production or high-yield of protein expression. Somehow, it is no doubt that Nglycosylation is involved in protein secretion.
In contrast to the Hex 8-9 GlcNAc 2 structures in T. reesei, the major glycoform N-glcan in A. fumigatus is Man 6 GlcNAc 2 [17]. In A. fumigatus, deletion of the Golgi α-mannosidase gene msdS causes defective N-glycan processing and gives rise to Man 8 GlcNAc 2 glycoform, which is similar with that of the wild-type T. reesei. Interestingly, the conversion of glycoform from Man 6 GlcNAc 2 to Man 8 GlcNAc 2 leads phenotypes such as defective cell wall, reduced conidiation and abnormal polarity in A.
fumigatus [17], suggesting a species-specific N-glycan structure in different filamentous fungi. To verify this hypothesis, in this study we introduced the A. fumigatus msdS gene into T. reesei.
The gpdA gene encodes glyceraldehyde-3-phosphate dehydrogenase (GPD) and is a key enzyme in glycolysis and glucogenesis which constitutes up to 5% of the soluble cellular protein in Saccharomyces cerevisiae [34] and A. nidulans [35]. Several copies of GPD-encoding genes have been accounted in higher eukaryotes [35][36], but only a single GPD-encoding gene has been reported in A. nidulans [37]. The gpdA-promoter-controlled exocellular production of glucose oxidase by recombinant A. niger NRRL-3 during growth on glucose and non-glucose carbon sources was investigated for better carbon substrate identification [38]. As several studies have successfully

Conclusion
Overall, although growth rate and cell wall integrity were not altered, phenotypes such as increased cell wall components, increased conidiation, abnormal polarity and enhanced cellulase activity were documented in strain Tr-MsdS. In conclusion, our results confirm that N-glycan processing between A. fumigatus and T. reesei is species-specific and required for protein secretion. In addition, our results also provide a new strategy to improve cellulases production by interfering the N-glycan processing in T. reesei.

Protein extraction
The extraction of proteins from different fractions of cells was done as described by Wang et al. [19] with some modifications. The extracellular proteins were precipitated with freshly prepared 2% (w/v) sodium deoxycholate to the culture supernatant (1:100) for 30 min on ice followed by addition of 100% trichloroacetic acid (1:10) for 30 min on ice, and collected by centrifugation (12,

Analysis of N-glycan
N-glycans were released from membrane and secreted proteins of T. reesei strain by peptide Nglycosidase F (PNGase F, NEB, P0704) [19]. The enzyme reaction includes the process of denaturation of proteins at 95°C for 5 min followed by addition 10% Nonidet P40 (NP40) treating with PNGase F.
After digestion, the sample was centrifuged and the supernatant was subjected to C8 column, washed with 100% acetonitrile (ACN) and equilibrated with 0.1% trifluoroacetic acid (TFA) to separate Nglycans from proteins. The released N-glycans were collected and applied to a graphite column, washed with 0.1% TFA to remove salts and then eluted with elution buffer (60% ACN, 0.1% TFA) to collect N-glycans. The structure of released and purified N-glycans was analyzed by MALDI-TOF-MS.

Phenotypic analysis
The growth kinetics was determined by measuring colony diameter. Same amount of spore was inoculated on media plate and the colony growth was monitored by measuring the diameter of each colony at different time intervals [19]. Similarly, the sensitivity of the mutant to antifungal reagents was also analyzed [17,19]. The conidiophores were spotted on solid PDA medium with uridine (MMU) plates with same concentration in the presence of 100 µg/ml calcofluor white, 150 µg Congo red or 40 µg SDS and incubated at 28°C, 32°C and 37°C and the plates were analyzed for colony growth, measured for colony diameter and photographed if required.

Electron Microscopy
Scanning and transmission electron microscopy analyses were performed as described by Li et al. [17]. Culture samples were fixed with 2.5% glutaraldehyde in phosphate buffer pH 7. Cells were embedded in mold at 55°C for 24 h followed by sectioning (Leica uc7) and staining (uranyl acetate for 25 min and lead citrate for 5 min) and the sections were examined with JEM-1400. Germination 10 6 freshly harvested conidia were inoculated in 10 ml of minimal liquid medium in a Petri dish containing 5-6 glass coverslips at 32°C. The coverslips with adhering germinated conidia were taken out and counted for the number of germ tubes germinated at the specified times counted under differential interference contrast microscopy [17].

Quantitative Real Time PCR
One hundred milliliters of complete liquid medium were inoculated with 10 6 -10 7 conidia and the harvested mycelia were disrupted by grinding. Total RNA was extracted using TRIzol reagent (Invitrogen/Life Technologies, Carlsbad, CA, US) [19,46]. The cDNA synthesis was carried out with 5 µg RNA using the RevertAid™ First Strand cDNA Synthesis Kit (Fermentas). The forward and reverse primers used for cDNA synthesis was MsdS-FGCTTGCTTTGATGGAGGAAG and MsdS-R TGACGCGGTACGCATAGTAG, respectively. The quantitative PCR reaction was done with SYBR®PremixExTaq™ (Takara) with the thermal cycling condition of 95°C for 30 s, followed by 40 cycles of 95°C for 5 s and 60°C for 30 s (Bio-Rad CFX manager 3.1). Quantification of mRNA levels of different genes was performed using the 2 -ΔΔct method. The 18s rRNA gene was used to standardize the mRNA levels of the target genes. Each assay and each experiment were repeated 3 times. To avoid or detect any possible contamination or carryover, appropriate negative controls containing no template were also subjected to the same procedure. Primers used in this study are listed in Table S3.

Enzyme activity assays and hydrolysis yield
The secreted proteins from different strains were analyzed for cellulase and b-mannanase activity as well as reducing sugar yield at three different condition of temperature (30°C, 40°C and 50°C) using carboxymethyl cellulose or locust bean gum as a substrate (Sigma-Aldrich Co.). The activity assay

Availability of data and materials
The datasets supporting the conclusions of this article are included within the article and its Additional file.

Ethics approval and consent to participate
Not applicable.

Consent for publication
Not applicable.     Phenotype and morphology of Tr-MsdS strain. In A, sensitivity to cell wall perturbing compounds was carried out by spotting conidiophores on solid MMU plates supplemented with 100 µg/ml calcofluor white, 150 µg Congo red or 40 µg SDS and incubating at 28°C, 32°C and 37°C. In B, the mycelia and conidial cells were fixed as described in method section and examined with transmission electron microscopy (TEM) (JEM-1400). In C, thickness (µm) of cell wall was measured under TEM.

Figure 4
Activation of the genes involved in protein transport in T. reesei. 106-107 conidia were inoculated 100 ml of complete liquid medium and cultured at 32 °C. Mycelia were harvested and disrupted by grinding. Total RNA was extracted and quantification of mRNA levels were performed as described under method section. The 18s rRNA gene was used to standardize the mRNA levels of the target genes. Primers used in this assay are listed in Table S3. Each assay and each experiment were repeated 3 times. Results were presented as mean SD.    and enzyme activity (β-mannanase or cellulase, respectively) (B) were measured. Results were presented as mean SD.

Supplementary Files
This is a list of supplementary files associated with this preprint. Click to download. Supplementaldata.docx