Characterisation of novel-cell-wall LysM-domain proteins LdpA and LdpB from the human pathogenic fungus Aspergillus fumigatus

Aspergillus fumigatus, a filamentous fungus that is ubiquitous in the environment, causes several human pulmonary disorders, including chronic and acute invasive infections and allergic diseases. Lysin motif (LysM) is a small protein domain that binds chitin, a major component of fungal cell wall polysaccharides. Several secreted LysM-domain proteins without catalytic function (LysM effectors) have been identified. They act as virulence factors in plant pathogenic fungi by preventing the immune response induced by chitin; however, LysM proteins in mammalian pathogenic fungi remain largely unexplored. We describe two novel LysM-domain proteins, LdpA and LdpB, in A. fumigatus. Functional analyses of single and double knockouts revealed no significant effects on cell wall chitin content, cell wall integrity, fungal morphology and fungal growth. Fluorescent signals from LdpA-green fluorescent protein (GFP) and LdpB-GFP were observed in cell wall and extracellular matrix. In a mouse model of invasive pulmonary aspergillosis, survival did not differ between ΔldpA/B and wild-type infection; however, further studies are required to reveal their functions in fungal−host interactions.

Lysin motif (LysM) was first described as a protein domain within the C-terminus of the lysozyme of bacteriophage 7 . Subsequent studies revealed that this motif is found in various proteins from prokaryotes and eukaryotes and bind polysaccharides, which contain N-acetylglucosamine (GlcNAc) residues including chitin and peptidoglycan 8 . Most bacterial LysM containing proteins are peptidoglycan hydrolases with various cleavage specificities 8 . In fungi, the LysM-domain is found predominantly in subgroup C chitinases 9 and LysM effectors, which are secreted proteins with multiple LysM domains but have no catalytic domain 10 . Several LysM effectors have already been identified as virulence factors in plant pathogenic fungi 11 . For instance, the tomato pathogen Cladosporium fulvum prevents chitin-triggered immunity by secreting the LysM effector Ecp6 12,13 . Similarly, in the rice blast fungus Magnaporthe grisea, LysM effector Slp1 suppresses chitin-induced plant immune responses 14 . In contrast to LysM effectors in plant pathogenic fungi, little is known about the expression and function of LysM proteins in mammalian pathogenic fungi.
Mammals do not synthesise chitin but produce enzymatically active chitinases such as chitotriosidase 15,16 and acidic mammalian chitinase 17,18 . Fungal cell wall chitin acts as a pathogen-associated molecular pattern and is reportedly a potential inducer of allergic inflammation 19 . Some reports have also indicated that the cell wall chitin of A. fumigatus recruits lung eosinophils [20][21][22][23] . These studies suggest that mammalian pathogenic fungi produce LysM effector proteins to circumvent chitin-triggered host immunity, similar to plant pathogens. Moreover, a recent whole-genome sequence analysis revealed that putative LysM effector proteins are widespread in the fungal kingdom, including mammalian pathogenic species 19 .
In this article, we identified novel LysM-domain protein A (LdpA) and B (LdpB) in A. fumigatus. We then investigated their protein functions by using single-gene deletion mutants ΔldpA and ΔldpB and double-gene www.nature.com/scientificreports www.nature.com/scientificreports/ LdpA and LdpB localise in the cell wall and extracellular matrix (eCM). To characterise the subcellular and extracellular distribution of LdpA and LdpB, A. fumigatus strains expressing LdpA-GFP fusion protein (AfS35-LdpA-GFP) and LdpB-GFP fusion protein (AfS35-LdpB-GFP) were generated (Fig. S3). Cell wall chitin was stained with CFW and was observed under a fluorescent microscope. Both LdpA-GFP and LdpB-GFP were visible in the hyphal cell wall ( Fig. 2A), conidial cell wall (Fig. 2B) and ECM (Fig. 2C). In the control mutant expressing unfused GFP (AfS35-GFP), fluorescence was observed in the cytoplasm but not in the cell wall ( Fig. 2A, B). To analyse LdpA and LdpB distribution in greater detail, ECM was stained with rhodamine-conjugated wheat germ agglutinin (WGA), which is a chitin-and sialylated glycans-binding lectin. Then ECM was observed under an invertible fluorescent microscope. No GFP fluorescence in the ECM was observed beneath AfS35-GFP cultures, but such signals were robust beneath colonies expressing LdpA-GFP or LdpB-GFP, thus confirming secretion into the ECM (Fig. 2D). LdpA-GFP and LdpB-GFP were observed in the ECM bound to the glass surface under a laser scanning confocal microscope (Fig. 2E). To further confirm the localisation of LdpA and LdpB, culture supernatants (CSs), hyphal cell wall fractions (CWFs) and hyphal cytosolic fractions (CFs) were isolated and subjected to Western blot analysis with an anti-GFP antibody (Ab). Protein bands corresponding to LdpA-GFP and LdpB-GFP were observed in hyphal CWFs but not in CSs and hyphal CFs (Fig. 2F). Both LdpB-GFP and GFP protein bands on SDS-PAGE were at their predicted molecular weights (MWs) (59 and 26 kDa, respectively) (Fig. 2F). However, the LdpA-GFP protein band on SDS-PAGE was approximately 20 kDa heavier than its predicted MW (64 kDa).

Discussion
Plant pathogenic fungi secrete various LysM effectors to interfere with recognition of chitin fragment by host immune 11 . However, such LysM effectors have not been identified in mammalian pathogenic fungi. In the present study, we describe the characteristics of two novel LysM-domain proteins, namely, LdpA and LdpB, from the human pathogenic fungus A. fumigatus. We also demonstrate that such proteins are secreted like LysM effectors; therefore, these proteins possibly influence currently unidentified host-pathogen interactions.
In fungi, the LysM-domain is predominantly found in subgroup C chitinases 9 and LysM effectors 10 . Pfam domain search revealed that A. fumigatus Af293 has eight putative LysM-domain proteins, including three putative chitinases (Afu5g03960, Afu5g06840 and Afu6g13720). LdpA (Afu5g03980) and LdpB (Afu1g15420) have multiple putative LysM domains but have no catalytic domain; this finding is consistent with LysM effectors. Interestingly, ldpA and ldpB was primarily expressed in hyphae and dormant conidia, respectively, suggesting that LdpB might have a function related to the conidial dormancy. Furthermore, the functional analysis of single-and double-deletion mutants revealed that LdpA and LdpB have no significant effects on fungal morphology, fungal growth, cell wall integrity or chitin contents in hyphae, thus suggesting that LdpA and LdpB are not essential for the biosynthesis of cell wall chitin and cell wall integrity under laboratory conditions. Many fungal cell wall proteins are heavily glycosylated and have a very high and variable apparent molecular mass when separated in gels 28 . The LdpA-GFP protein band on SDS-PAGE was approximately 20 kDa heavier than its predicted MW, thus suggesting that LdpA could be modified by posttranslational glycosylation.
The plant pathogenic fungus C. fulvum has a chitin-binding LysM effector, namely, Ecp6, with high affinity for various short-chain chitin oligosaccharides. This binding acts to prevent the activation of chitin-triggered immunity 12,13 . LdpA-GFP or LdpB-GFP fluorescence signals were found in the cell wall ECM and colocalised with cell wall chitin as revealed by CFW and WGA staining, thus further suggesting that LdpA and LdpB could be chitin-binding LysM effectors. However, in an artificial infection experiment using the mouse IPA model, survival did not differ significantly between mice infected with ΔldpA/B or WT strains, thus suggesting that LdpA and LdpB are not virulence factors of A. fumigatus. Nonetheless, these proteins may still influence host interactions. A. fumigatus forms multicellular communities in vitro and in vivo (i.e. termed biofilms) that are composed of hyphae and ECM 29,30 and promote antifungal drug resistance 31 . We demonstrated that the ECM contained both proteins. Therefore, LdpA and LdpB could influence biofilm formation and antifungal susceptibility. These mutants could be also useful tools for visualising the ECM, investigating the mechanism of biofilm formation and elucidating the ultimate functions of these proteins. Although secreted LysM proteins are strongly implicated in pathogenesis, numerous additional functions are assumed on the basis of the variety of niches colonised by these fungi.
In conclusion, we describe the LdpA and LdpB of the human pathogenic fungus A. fumigatus. We demonstrate that LdpA and LdpB are localised to the cell wall and ECM; however, they have no capacity to influence the morphology or acute pathogenicity of A. fumigatus.

Methods
Fungal strains. Table 1 shows the strains used in this study. A. fumigatus strains were maintained on PDA (BD Biosciences) at 25 °C. In all experiments, A. fumigatus conidia were prepared as follows: after 5-7 days of culture on PDA at 35 °C, conidia were harvested with 3G3 glass filters (AGC Techno Glass) by using 0.05% Tween 20, resuspended in phosphate buffer saline (PBS) with 0.02% Tween 20 and counted using a haemocytometer. Generation of ΔldpA, ΔldpB and ΔldpA/B. Gene disruption constructs were generated according to the methods described by Higuchi et al. 32 and Kuwayama et al. 33 To generate the ldpA single-gene deletion mutant (ΔldpA) and ldpB single-gene deletion mutant (ΔldpB), approximately 1 kbp of 5′-flanking region and 3′-flanking region were PCR amplified from A. fumigatus AfS35 genomic DNA. The hygromycin B resistance gene (hph) cassette was PCR amplified from pBC-hygro (Fungal Genetics Stock Center). The PCR fragments were fused by overlap extension PCR. The resulting gene replacement constructs were used for the transformation of A. fumigatus AfS35 by the polyethylene glycol (PEG)-mediated protoplast transformation method 34 (Fig. S1). Transformants were selected for growth in the presence of hygromycin B. To generate the double-gene deletion mutant ΔldpA/B, approximately 1 kbp of ldpA 5′-flanking region and 3′-flanking region of ldpA were amplified from A. fumigatus AfS35 genomic DNA. The pyrithiamine resistance gene (ptrA) cassette was PCR amplified from pPTRII (Takara Bio). The PCR fragments were fused by overlap extension PCR. The resulting gene replacement construct was used for the transformation of A. fumigatus ΔldpB by the PEG-mediated protoplast transformation method (Fig. S2). Transformants were selected for growth in the presence of pyrithiamine.
PCR was performed using high fidelity DNA polymerases (PrimeSTAR ® DNA Polymerase; Takara Bio). Gene deletion was confirmed by PCR and real-time quantitative PCR (Figs S1 and S2). Table S1 shows the plasmids used in this study.

Generation of A. fumigatus expressing LdpA-GFp or LdpB-GFp fusion protein. Complete ldpA
and ldpB CDSs were amplified from pCR2.1-LdpA and pCR2.1-LdpB. The PCR products were inserted into pHAN02-GFP (Fig. S3) between the HindIII and SmaI sites by using the In-Fusion HD Cloning Kit (Takara Bio). The resulting plasmids were transformed into chemically competent E. coli strain HST08. The plasmids were extracted using the GenElute TM Plasmid Miniprep kit (Sigma-Aldrich) and linearised using the restriction enzymes BamHI or EcoRI. The A. fumigatus niaD − mutant AfS35-niaD − was isolated by positive selection using chlorate according to the methods described by Unkles et al. 35 and Ishi et al. 36 . The resulting plasmids were used for the transformation of AfS35-niaD − by PEG-mediated protoplast transformation methods. The niaD +

Strains
Described in this study Relevant characteristics Source www.nature.com/scientificreports www.nature.com/scientificreports/ revertants were selected for growth on Czapek-Dox agar (Oxoid), which contains 1.2 M sorbitol and sodium nitrate as the sole source of nitrogen. Gene integration was confirmed by PCR.
WGA staining of eCM. After 72 h of stationary culture on glass bottom dishes (AGC Techno Glass) that contain RPMI 1640 medium at 37 °C under 5% CO 2 , adherent fungal communities were gently washed three times with PBS and then incubated with 10 µg/mL rhodamine-conjugated WGA (Vector Laboratories) for 15 min at room temperature. After washing three times with PBS, ECM formation beneath the colonies was observed under an inverted fluorescent microscope. Measurement of chitin content. Hyphal cell wall chitin was measured as described by Tomishige et al. 35,37 with some modifications. In the current study, 100 mg of hyphae cultured in AMM liquid medium was resuspended in 1 mL of 6% KOH and incubated at 80 °C for 90 min. After cooling at room temperature, 100 μL of glacial acetic acid was added. Insoluble materials were washed twice with water, resuspended in 600 μL of 50 mM potassium phosphate (pH 7.5) containing 1 U Pyrococcus furiosus thermostable chitinase (Wako) and incubated at 85 °C for 2 h. After cooling at room temperature, 0.25 mg of Helix pomatia β-glucuronidase (Sigma-Aldrich) was added and incubated at 37 °C for 1 h. An aliquot of the mixture was assayed for GlcNAc content according to the procedure described by Reissig et al. 38 .

Subcellular localisation of
Fluorescence microscopy imaging. An upright fluorescent microscope (AXIO Imager A1, Carl Zeiss) with Zeiss filter sets 38 HE and 49 were used to observe GFP fluorescence and CFW fluorescence, respectively. An inverted fluorescent microscope (BZ-9000, Keyence) was used to observe the ECM formed at the bottom of glass dishes with a BZ filter GFP for observing GFP fluorescence and a BZ filter TRITC for observing rhodamine fluorescence. A laser scanning confocal microscope (LSM5 EXCITER, Carl Zeiss) was used to observe the GFP fluorescence in the ECM that formed at the bottom of the glass dishes.
Animals. Specific pathogen-free male ICR mice aged 5-6 weeks were purchased from Charles River Laboratories, Japan. All animal experiments were approved by the Committee on Animal Experiments of Chiba University and carried out according to the Chiba University Animal Experimentation Regulations.
Neutropenic mouse model of IPA. ICR mice were immunosuppressed by the intraperitoneal administration of 150 mg/kg body weight cyclophosphamide 1 and 4 days before infection and 3, 6, 9 and 12 days after infection and by the subcutaneous administration of 200 mg/kg hydrocortisone acetate 1 day before infection. Mice were infected by the intranasal administration of 30 μL PBS with 0.05% Tween20 (PBST) containing 3 × 10 4 conidia obtained from ΔldpA/B or WT A. fumigatus. Control mice were administered PBST without conidia. After infection, the mice were observed every day up to day 14. statistics. All statistical analysis was performed using GraphPad InStat 3 software. One-way ANOVA with post-hoc Tukey-Kramer tests were used to assess the statistical significance. A P value < 0.05 was considered significant for all tests.

Data Availability
The datasets during and/or analysed during the current study available from the corresponding author on reasonable request.