Discovery of α-(1→6)-linked mannan structures resembling yeast N-glycan outer chains in Aspergillus fumigatus mycelium

ABSTRACT The cellular surface of the pathogenic filamentous fungus Aspergillus fumigatus is enveloped in a mannose layer, featuring well-established fungal-type galactomannan and O-mannose-type galactomannan. This study reports the discovery of cell wall component in A. fumigatus mycelium, which resembles N-glycan outer chains found in yeast. The glycosyltransferases involved in its biosynthesis in A. fumigatus were identified, with a focus on two key α-(1→2)-mannosyltransferases, Mnn2 and Mnn5, and two α-(1→6)-mannosyltransferases, Mnn9 and Van1. In vitro examination revealed the roles of recombinant Mnn2 and Mnn5 in transferring α-(1→2)-mannosyl residues. Proton nuclear magnetic resonance (1H-NMR) analysis of cell wall extracts from the ∆mnn2∆mnn5 strain indicated the existence of an α-(1→6)-linked mannan backbone in the A. fumigatus mycelium, with Mnn2 and Mnn5 adding α-(1→2)-mannosyl residues to this backbone. The α-(1→6)-linked mannan backbone was absent in strains where mnn9 or van1 was disrupted in the parental ∆mnn2∆mnn5 strain in A. fumigatus. Mnn9 and Van1 functioned as α-(1→6)-linked mannan polymerases in heterodimers when co-expressed in Escherichia coli, indicating their crucial role in biosynthesizing the α-(1→6)-linked mannan backbone. Disruptions of these mannosyltransferases did not affect fungal-type galactomannan biosynthesis. This study provides insights into the complexity of fungal cell wall architecture and a better understanding of mannan biosynthesis in A. fumigatus. IMPORTANCE This study unravels the complexities of mannan biosynthesis in A. fumigatus, a key area for antifungal drug discovery. It reveals the presence of α-(1→6)-linked mannan structures resembling yeast N-glycan outer chains in A. fumigatus mycelium, offering fresh insights into the fungal cell wall’s design. Key enzymes, Mnn2, Mnn5, Mnn9, and Van1, are instrumental in this process, with Mnn2 and Mnn5 adding specific mannose residues and Mnn9 and Van1 assembling the α-(1→6)-linked mannan structures. Although fungal-type galactomannan’s presence in the cell wall is known, the existence of an α-(1→6)-linked mannan adds a new dimension to our understanding. This intricate web of mannan biosynthesis opens avenues for further exploration and enhances our understanding of fungal cell wall dynamics, paving the way for targeted drug development.

FTGM α-core mannan synthesis, leads to abnormal colony morphology and a reduction in infectious capacity in A. fumigatus (6,7).Furthermore, the double disruption of pmt4 and pmt1, encoding protein O-mannosyltransferases, is synthetically lethal in A. fumigatus, emphasizing the essential role of protein O-mannosylation in hyphal growth through the maintenance of cell surface and secreted proteins (8).Additionally, disrupting mnt1, an α-(1→2)-mannosyltransferase gene responsible for decorating the second mannosyl residue to protein O-mannose-type galactomannan (OMGM) and elongating high mannose type N-glycan core chain, results in a thin cell wall and hypovirulence in A. fumigatus (9,10).Therefore, comprehensively understanding mannan biosynthesis in A. fumigatus is crucial for advancing novel antifungal drug discovery.
In our study, we specifically focused on two putative α-(1→2)-mannosyltransferase, Mnn2 and Mnn5, in A. fumigatus.Recombinant AfMnn2 and AfMnn5 had α-(1→2)mannosyltransferase activity in vitro.Although ∆mnn2 or ∆mnn5 mutants showed no discernible phenotype, the ∆mnn2∆mnn5 double mutant exhibited a growth defect and abnormal conidial formations in A. fumigatus.Proton nuclear magnetic resonance ( 1 H-NMR) analysis of the total mannan fraction from the ∆mnn2∆mnn5 mutant provi ded compelling evidence for the presence of α-(1→6)-linked mannan backbone in A. fumigatus mycelia.Furthermore, Mnn9 and Van1 were identified as key enzymes in the biosynthesis of this mannan backbone in A. fumigatus.In our in vitro experiments, recombinant AfMnn9 and AfVan1 were shown to form a heterodimer and catalyze the synthesis of α-(1→6)-linked mannose polymers.

Mannosyltransferase activities of Mnn2 and Mnn5 in vitro
In our investigation of the mannosyltransferase activities of Mnn2 (AFUB_093840) and Mnn5 (AFUB_060800) from A. fumigatus, we produced individual N-terminal 6×Histagged recombinant proteins using a bacterial expression system.Mnn2 and Mnn5 were expressed without the predicted transmembrane domains, spanning amino acid residues 1-30 and 1-29, respectively.Analysis of the purified recombinant proteins through sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) revealed bands close to their predicted molecular weights of 53 and 52 kDa, respectively (Fig. 1A).Subsequently, we assessed mannosyltransferase activity at 30°C for 16 h using recombinant Mnn2 and Mnn5 (0.1 µg/µL each).The reaction involved p-nitrophenyl α-D-mannopyranoside (α-Man-pNP, 1.5 mM) as the acceptor substrate, guanosine diphosphate-α-D-mannose (GDP-Man, 10 mM) as the donor substrate, and 0.5 mM Mn 2+ as the metal cofactor.Substrates and products were analyzed using reverse-phasehigh-performance liquid chromatography (HPLC).Fractions with heat-inactivated Mnn2 and Mnn5 exhibited no new peaks (Fig. 1B, upper panels).Conversely, the fractions containing Mnn2 or Mnn5 revealed new products (referred to as Mnn2-product and Mnn5-product) at 19.5 min (Fig. 1B, middle and bottom panels).To elucidate the chemical structures of Mnn2-product and Mnn5-product, we isolated each peak and digested them using a substrate-specific mannosidase.These products were analyzed by normal-phase HPLC.Recombinant MsdS/MsdC, an α-(1→2)-specific mannosidase from A. fumigatus, was employed to convert Mnn2-product and Mnn5-product to α-Man-pNP (Fig. 1C).These findings provide compelling evidence supporting the role of Mnn2 and Mnn5 as α-(1→2)-mannosyltransferases.
ΔScmnn2ΔScmnn5, respectively, resulted in the appearance of chemical shifts at 5.04 and 5.14 ppm, although the chemical shift at 4.9 ppm was not completely eliminated (Fig. 4).The result indicates that AfMnn2 and AfMnn5 partially complement the function of the budding yeast ScMnn2p.These results suggest that the functions of AfMnn2 and AfMnn5 are the same as ScMnn2p and that they do not, or do very weakly, facilitate transfer of the second mannosyl residue.

Structural analysis of α-(1→6)-linked mannan from A. fumigatus wild-type, Δmnn2Δmnn5, Δmnn2Δmnn5Δmnn9, Δmnn2Δmnn5Δvan1, and Δmnn2Δmnn5ΔanpA strains
To identify the glycosyltransferases involved in the biosynthesis of A. fumigatus, we extracted the total mannan from wild-type, Δmnn2Δmnn5, Δmnn2Δmnn5Δmnn9, Δmnn2Δmnn5Δvan1, and Δmnn2Δmnn5ΔanpA strains and analyzed their structures (Fig. 5).The disruption of mnn9 or van1 in the Δmnn2Δmnn5 strain resulted in the disappear ance of the 4.9 ppm chemical shift observed in Δmnn2Δmnn5 (Fig. 5).This suggests that Mnn9 or Van1 are essential for the biosynthesis of α-(1→6)-linked mannan in A. fumiga tus.In the Δmnn2Δmnn5ΔanpA strain, the signal of FTGM was absent, but the signal of α-(1→6)-linked mannan was unaffected.This indicates that AnpA is involved only in the biosynthesis of FTGM, but not α-(1→6)-linked mannan in A. fumigatus (Fig. 5).

Is α-(1→6)-linked mannan bound to the outer chain of N-glycan?
Next, we evaluated the involvement of Mnn2, Mnn5, Mnn9, and Van1 in N-glycan biosynthesis in A. fumigatus using a N-glycosylated protein, SucA.The SucA was detected by immunoblotting to estimate the length of N-glycan.If these mannosyltransferases contribute to N-glycan elongation, a lower apparent molecular weight should be observed in these strains than in the A1151 strain, as SucA has nine potential N-gly cosylation sites.SucA was detected as two major bands, including one high-and one low-molecular-weight band.Both main bands had increased mobility following endoglycosidase (Endo) H treatment, indicating that these SucA proteins were N-gly cosylated (Fig. S4, lanes 1 and 5).The apparent molecular weight of both types of SucA expressed in ∆mnn2∆mnn5, ∆mnn9, and ∆van1 strains was comparable with SucA expressed in A1151 (Fig. S4, lanes 1-4).This suggests that the α-(1→6)-linked mannan is derived from a glycan different from the N-glycan outer chain or a short N-glycan outer chain structure.

DISCUSSION
In this investigation, we delved into the functional roles of Mnn2 and Mnn5, contributors to hyphal growth and conidial formation through the biosynthesis of mannan structures in A. fumigatus.
The 1 H-NMR analysis of the total mannan fraction unveiled the presence of α-(1→6)linked mannan structures in A. fumigatus mycelium.The distinctive chemical shift originating from the α-(1→6)-linked mannan becomes discernible when the side chain is absent (22,(32)(33)(34).Both AfMnn2 and AfMnn5 emerge as pivotal catalysts in side-chain biosynthesis, indicating their collaborative role in transferring the initial α-(1→2)mannosyl residues to the α-(1→6)-linked mannan backbone (Fig. 9).In S. cerevisiae, ScMnn5p is recognized for transferring the second α-(1→2)-mannosyl residue of the side chain (18).However, upon heterologous expression of AfMnn2 or AfMnn5, the mannan (Continued on next page) structure of ΔScmnn2ΔScmnn5 was restored to that of ΔScmnn5 (Fig. 4), indicating that AfMnn2 and AfMnn5 share the same function as ScMnn2p; however, they do not have the enzymatic function of ScMnn5p.This suggests that the chain length of the α-(1→2)mannosyl side chain of α-(1→6)-linked mannan may be limited to one in A. fumigatus (Fig. 9).The detailed structural analysis of this glycan in A. fumigatus could clarify this conclusion.We uncovered the involvement of Mnn9 and Van1, members of the GT62 family, in mycelial α-(1→6)-linked mannan biosynthesis (Fig. 5).The recombinant heterodimer of Mnn9 and Van1 was found to function as a mannan polymerase in vitro (Fig. 6 to 8).Contrary to prior investigations documenting Mnn9's manifestation of mannosyltransferase activity in isolation (28), our results suggest that neither Mnn9 nor Van1 has mannosyltransferase activity independently in vitro (Fig. 7).The present study used lower enzyme concentrations, lower substrate and substrate manganese concen trations, and different reaction conditions (e.g., omission of DTT) than Henry et al, which could account for the differences in recorded enzyme activity (28).However, the homologous GT62-family protein in A. fumigatus, AnpA, exhibits strong mannosyltrans ferase activity alone under the reaction conditions used in our previous study (35).Henry et al. may have observed very weak activity (28).Furthermore, our results align with the obliteration of the α-(1→6)-linked mannan-derived chemical shift following the singular disruption of either mnn9 or van1 (Fig. 5).These findings underscore the essential collaborative role of both mnn9 and van1 in the biosynthesis of the α-(1→6)-linked mannan in A. fumigatus.In the context of A. fumigatus, the mnn2 and mnn5 doubledisruption strain displayed a reduced mycelial elongation rate and diminished ability to form conidiophores (Fig. 2).Notably, reports indicate that Δmnn9 and Δvan1 do not exhibit the Δmnn2Δmnn5-like phenotype (28,35).This discrepancy suggests that Mnn2 and Mnn5 may also be involved in the biosynthesis of mannans beyond the α-(1→6)linked mannan side chains.Mnn2 and Mnn5 are potentially involved in the biosynthesis of mannosyl residues in A. fumigatus glycolipids.However, A. fumigatus strains deficient in mannosyltransferase or N-acetylglucosaminyltransferase to inositol phosphorylcera mide exhibit normal mycelial growth (36)(37)(38), suggesting that Mnn2 and Mnn5 are unlikely to be involved in glycolipid biosynthesis.The intricate landscape of mannan biosynthesis in A. fumigatus remains complex and warrants further elucidation in future studies.
Mnn9 and Van1 in A. fumigatus have undergone extensive scrutiny, with recent findings linking them to the biosynthesis of the conidia-specific G3Man structure (30).Our study contributes a novel revelation by showcasing, for the first time, the presence of α-(1→6)-linked mannan in both mycelia and conidia.Intriguingly, Mnn9 and Van1 play a role in the biosynthesis of α-(1→6)-linked mannan in mycelial and conidial forms (30).However, the structural disparities in the side chains between conidial and mycelial α-(1→6)-linked mannans suggest that, in comparison to the complex G3Man, the mycelial α-(1→6)-linked mannan exhibits a simpler structure (Fig. 9).This structural distinction is likely attributed to the necessity for a complex structure, such as G3Man, in conidia due to the high percentage of β-glucan and the absence of galactosaminoga lactan (GAG) (30).Conversely, the presence of α-glucan and GAG in the mycelium may render a simpler mannan structure sufficient.The presence of mannans with different side-chain structures in the cell walls of conidia and mycelia remains unparalleled among known polysaccharides.Therefore, investigating the specific functional roles of these mannans in conidia and mycelia presents an intellectually compelling avenue for research.The physiological significance of these mannan structures warrants further elucidation.
To determine whether the α-(1→6)-linked mannan structure of A. fumigatus is part of the N-glycan, we investigated the changes in the apparent molecular weight of the invertase SucA on SDS-PAGE (Fig. S4).Loss of outer chains of N-glycans in budding and fission yeasts dramatically reduced the apparent molecular weight of N-glycosylated proteins on SDS-PAGE, such as invertase and acid phosphatase (12,39,40).However, in A. fumigatus, the apparent molecular weight change of SucA upon Endo H treatment was smaller than that in yeast (Fig. S4, lanes 1 and 5), consistent with those reported by other groups (28,29).These results suggest that the N-glycans of A. fumigatus do not possess the long outer chain structure present in the yeast.Moreover, we observed no remarkable change in the mobility of SucA in the Δmnn2Δmnn5, ∆mnn9, and ∆van1 strains compared with that in A1151 (Fig. S4, lanes 1-4).This may be because the N-glycosylated proteins of A. fumigatus do not have an N-glycan outer chain or as the N-glycans of A. fumigatus are not as long as those of yeast.In this paper, no conclusions could be drawn as to what the α-(1→6)-linked mannan structure is bound to.Du et al. proposed that α-mannan elongation occurs not via N-glycan but through O-linked mannose in cell wall mannoproteins (29).Unlike yeasts, which heavily rely on outer chain structures for cell wall integrity, the presence of other mannan structures, such as FTGM, in A. fumigatus may diminish the importance of outer chain structures.Conversely, certain species, such as N. crassa, emphasize the significance of outer chain structure even in filamentous fungi, showcasing the intriguing diversity of mannan structures, including the outer chain, within the phylum Ascomycota (25).In filamentous fungi, different mannans are present in the cell wall.Because the structures of the mannans are similar, distinguishing between and analyzing them is difficult.Further studies are needed to clarify their differences.
In summary, our investigations into Mnn2 and Mnn5 have unveiled the presence of an α-(1→6)-linked mannan structure in the mycelia of A. fumigatus.These findings mark a substantial contribution to our comprehension of the structure and biosynthe sis of filamentous fungal cell wall, with potential implications for the development of antifungal and agrochemical agents.

Strains and medium
The A. fumigatus strains utilized in this investigation are detailed in Table S1 (Table S1), with A. fumigatus A1160 and A1151 strains serving as the parental and control strains, respectively.
The S. cerevisiae strains employed in this study are enumerated in Table S1.BY4741 and Δmnn5 strains were procured from the Yeast Knockout (YKO) Collection (Horizon Discovery Ltd., UK).Growth for these strains occurred on yeast extract-peptone-dextrose (YPD) medium or synthetic complete (SC) medium.

Construction of mnn2 and mnn5 gene disruption strains
A. fumigatus mnn2 or mnn5 mutants were made by replacing mnn2 with mnn2Δ::AnpyrG or mnn5Δ::AnpyrG cassettes.A gene replacement cassette, consisting of the homology arm at the 5′ end of the mnn2 and mnn5 genes and the homology arm at the 3′ end of the mnn2 and mnn5 genes, was generated by recombinant polymerase chain reaction (PCR).The A. fumigatus A1151 genomic DNA served as the template, and the primer pairs xxxx-1/xxxx-2 and xxxx-3/xxxx-4 were used for amplification (where "xxxx" indicates mnn2 or mnn5) (Table S2).Simultaneously, the AnpyrG marker was amplified by recombinant PCR using pHSG396-AnpyrG (35) as the template and the primer pair pHSG396-F/pHSG396-R.The resulting DNA fragment, amplified with primers xxxx-1 and xxxx-4, was employed for the transformation of A. fumigatus A1160, yielding Δmnn2 and Δmnn5 strains.MM agar plates lacking uracil and uridine were utilized for the selection of transformants.The successful introduction of AnpyrG into each gene locus was confirmed through PCR, using the primer pairs mnn2-F/pyrG-R and pyrG-F/mnn2-R, as well as mnn5-F/pyrG-R and pyrG-F/mnn5-R, respectively (Fig. S2A and B).

Construction of ∆mnn2∆mnn5 strains
A. fumigatus ∆mnn2∆mnn5 mutants were made by replacing mnn5 with mnn5Δ::ptrA or mnn5Δ::hph cassettes in A. fumigatus ∆mnn2.A gene replacement cassette, includ ing the homology arm at the 5′ end of the mnn5 gene and the homology arm at the 3′ end of the mnn5 gene, was generated by recombinant PCR using A. fumiga tus A1151 genomic DNA as the template and the primer pairs mnn5-1/mnn5-2 and mnn5-3/mnn5-4, respectively (Table S2).The ptrA and hph markers were amplified by recombinant PCR using pHSG396-ptrA and pHSG396-hph (10) as templates and the primer pair pHSG396-F/pHSG396-R.The resulting DNA fragments, amplified with primers mnn5-1 and mnn5-4, were used to transform A. fumigatus Δmnn2, generating Δmnn2Δmnn5 (ptrA) and Δmnn2Δmnn5 (hph), respectively.MM agar plates supplemen ted with 0.1 µL/mL pyrithiamine or 200 µL/mL hygromycin B were employed for the selection of transformants.The introduction of ptrA or hph into each gene locus was confirmed by PCR using the primer pairs mnn5-F/ptrA-R and ptrA-F/mnn5-R, mnn5-F/ hph-R, and hph-F/mnn5-R, respectively (Fig. S2C and D).

Construction of Δmnn2Δmnn5Δmnn9, Δmnn2Δmnn5Δvan1, and Δmnn2Δmnn5ΔanpA strains
The mnn9, van1, or anpA were disrupted in A. fumigatus Δmnn2Δmnn5 (ptrA) by hph insertion.A gene replacement cassette, including the homology arm at the 5′ end of the mnn9, van1, or anpA genes and the homology arm at the 3′ end of the mnn9, van1, or anpA genes, was generated by recombinant PCR using A. fumigatus A1151 genomic DNA as the template and the primer pairs xxxx-1/xxxx-2 and xxxx-3/ xxxx-4, respectively (where "xxxx" indicates mnn9, van1, or anpA) (Table S2).The hph marker was amplified by recombinant PCR using pHSG396-hph (10) as the template and the primer pairs pHSG396-F/pHSG396-R.The resulting DNA fragment, amplified with primers xxxx-1 and xxxx-4, was used to transform A. fumigatus A1160, generating Δmnn2Δmnn5Δmnn9, Δmnn2Δmnn5Δvan1, or Δmnn2Δmnn5ΔanpA strains, respectively.MM agar plates supplemented with 200 µL/mL hygromycin B were employed for the selection of transformants.The introduction of hph into each gene locus was confirmed by PCR using the primer pair xxxx-F/hph-R and hph-F/xxxx-R (Fig. S2E thorugh G).

Determination of colony growth rate
The colony growth rates were measured as described previously (43).Briefly, conidia from each strain were point-inoculated into the center of MM plates.The colony diameters were measured after 24, 48, 72, 96, and 120 h of incubation at 37°C, and growth rates in mm/h were determined at each incubation interval (i.e., 24-48, 48-72, 72-96, and 96-120 h).The rates were averaged across the entire time interval.Measure ments were performed 12 times for each strain.

Analysis of conidiation efficiency
The conidiation efficiency was analyzed as described previously (44), with slight modification.Briefly, conidia from each strain were point-inoculated into the center of MM plates.After 5 days of incubation at 37°C, the conidia were suspended in 5 mL of 0.01% (wt/vol) Tween 20 and counted using a hemocytometer.

Protein purification and quantification
Bacterial expression and purification of 6×His-tagged proteins were conducted following a previously described method (42).The yields of the recombinant proteins were 0.88 mg/L for Mnn2 and 1.25 mg/L for Mnn5.

Preparation of total mannan fraction from A. fumigatus
The total mannan fractions from A. fumigatus were prepared as follows.The total mannan was extracted from A. fumigatus mycelia by autoclaving at 121°C for 2 h with 100 mM citrate buffer (pH 7.0) (6,31,42,46).The cell wall extract was purified by cetyltrimethylammonium bromide (CTAB) using a previously described method (47), with slight modifications.CTAB (4 g) was dissolved in 150 mL of the cell wall extract and allowed to stand at 25°C for several hours.The solution was centrifuged at 12,000 × g for 15 min, and the supernatant was transferred to another tube.The precipitate was washed with 50 mL dH 2 O, followed by centrifugation at 12,000 × g for 15 min.The supernatant was combined with the previous supernatant.Boric acid was added at a concentration of 1% (wt/vol); while stirring, 2 M NaOH solution was added to adjust the pH to 8.8.The mixture was stirred well and stood for 1 h before centrifugation at 12,000 × g for 15 min.The precipitate was washed with 0.5% sodium acetate (pH 8.8) and dissolved in 50 mL of 2% acetic acid solution.Sodium acetate (1 g) and 150 mL of ethanol were added, followed by centrifugation at 12,000 × g for 15 min.The resultant sample was dialyzed in dH 2 O and lyophilized.To remove O-linked glycans, a β-elimination reaction was executed by subjecting the fractionated galactomannan to reduce alkali conditions (500 mM NaBH 4 /100 mM NaOH, 10 mL, at 25°C for 24 h).Following neutralization with a 50% acetic acid solution, the samples underwent overnight dialysis against distilled water.The purified samples were then lyophilized, resuspended in distilled water, and clarified using 0.45 µm pore filters.For the removal of galactofuran side chains, galactomannan underwent treatment with 100 mM hydrochloric acid at 100°C for 60 min (46).Subsequently, the samples were neutralized with 10 M NaOH and subjected to overnight dialysis against dH 2 O.The final product was defined as the total mannan fraction (46).

Preparation of mannoproteins from S. cerevisiae
Mannoproteins were extracted from S. cerevisiae cells by autoclaving at 121°C for 2 h in 100 mM citrate buffer (pH 7.0) (6,31,42,46).The extract was purified by CTAB using a previously described above.

H-NMR analysis
1 H-NMR analysis was performed following a previously established method (3,31).In preparation for nuclear magnetic resonance spectroscopy, samples for NMR were exchanged twice in D 2 O with intervening lyophilization.They were then dissolved in D 2 O (99.97% atom 2H).

Expression and detection of SucA
pPTR-II-SucA-a plasmid for expressing 3×FLAG-tagged SucA-was introduced into the A. fumigatus wild-type (A1151), Δmnn2Δmnn5 (hph), ∆mnn9, and ∆van1 strains (10), yielding A1151 +pPTR-II-SucA, ∆mnn2Δmnn5 + pPTR-II-SucA, ∆mnn9 +pPTR-II-SucA, and ∆van1 +pPTR-II-SucA, respectively.Each A. fumigatus mycelium was suspended in 1× glycoprotein denaturing buffer (0.5% SDS, 40 mM DTT) supplied with the Endo H product (New England Biolabs, Ipswich, MA, USA).Stainless steel beads were added to the samples, and the mycelia were crushed using a μT-12 bead cutter (TAITEC CORPO RATION, Koshigaya, Japan).After crushing, samples were centrifuged at 12,000 × g for 10 min, and the supernatant was transferred to another tube as the cell wall protein fraction.In the case of Endo H treatment, GlycoBuffer 3 (50 mM sodium acetate, pH 6.0), which was supplied with the Endo H product, and Endo H were added to the cell wall protein fraction, and the mixture was incubated at 37°C for 10 h.Cell wall protein fractions were used as samples for SDS-PAGE by adding SDS sample buffer and boiling the mixture for 10 min.Western blotting was performed according to the established method (31), and a 5,000-fold dilution of monoclonal anti-FLAG antibody (clone M2; Merck KGaA, Darmstadt, Germany) was used as the antibody.EzWestLumiOne (ATTO CORPORATION, Tokyo, Japan) was used for chemiluminescence detection.The chemiluminescence imager used was MicroChemi (Berthold Technologies, Bad Wildbad, Germany).

FIG 1
FIG 1 Mannosyltransferase activities of Mnn2 and Mnn5 in vitro.(A) SDS-PAGE analysis of purified recombinant Mnn2 and Mnn5.The purified recombinant proteins were separated by SDS-PAGE using a 5-20% gradient polyacrylamide gel.(B) Chromatograms of Mnn2 and Mnn5 mannosyltransferase activity assays using p-nitrophenyl α-D-mannopyranoside as the acceptor substrate.The 40 µL reaction mixture containing 50 mM HEPES-NaOH (pH 6.8), 100 mM NaCl, (Continued on next page)