rich peptidogalactomannan of Fusarium oxysporum is ...

Medical Mycology, 2018, 0, 1–12 doi: 10.1093/mmy/myx167 Advance Access Publication Date: 0 2018 Original Article

Original Article

β–1,6-linked Galactofuranose- rich peptidogalactomannan of Fusarium oxysporum is important in the activation of macrophage mechanisms and as a potential diagnostic antigen Nathalia Ferreira de Oliveira1 , Gustavo R. C. Santos2 , Mariana Ingrid D. S. Xisto1 , Giulia Maria Pires dos Santos3 , Marcio Nucci4 , Rosa Maria T. Haido5 and Eliana Barreto-Bergter1,∗ 1

´ ˆ ´ Instituto de Microbiologia Paulo de Goes, Centro de Ciencias da Saude, Universidade Federal do Rio ´ de Tecido ˜ 21941–970, Rio de Janeiro, RJ, Brazil, 2 Laboratorio de Janeiro (UFRJ), Bloco I, Ilha do Fundao, ´ Clementino Fraga Filho and Instituto de Bioqu´ımica Medica ´ Conjuntivo, Hospital Universitario Leopoldo de Meis, Universidade Federal do Rio de Janeiro (UFRJ), 21941-913, Rio de Janeiro, RJ, Brazil, 3 Instituto ´ ´ ´ RJ, Brazil, 4 Hospital Universitario Biomedico, Universidade Federal Fluminense (UFF), 24210-130, Niteroi, Clementino Fraga Filho, Universidade Federal do Rio de Janeiro (UFRJ), 21941-913, Rio de Janeiro, RJ, ´ UNIRIO, 20211-010, Rio de Janeiro, RJ, Brazil Brazil and 5 Instituto Biomedico,

∗

To whom correspondence should be addressed. Eliana Barreto-Bergter, PhD, Departamento de Microbiologia Geral, Instituto de Microbiologia da Universidade Federal do Rio de Janeiro (UFRJ), Rio de Janeiro, RJ, 21941–970, Brazil. Tel: 55 21–39386741; E-mail: [email protected] Received 2 October 2017; Revised 12 December 2017; Editorial Decision 19 December 2017

Abstract A peptidogalactomannan (PGM) from Fusarium oxysporum was structurally characterized by a combination of chemical and spectroscopic methods, including one and twodimensional nuclear magnetic resonance (1D and 2D NMR). The galactomannan component consists of a main chain containing (1→6)-linked β-D-galactofuranose residues with side chains containing (1→2)-linked α-D-Glcp, (1→2)-linked -β-D-Manp (1→2) and β-D-Manp terminal nonreducing end units and differs from that of Aspergillus fumigatus and Cladosporium resinae that present a main chain containing (1→6)-linked α-D-Manp residues presenting β-D-Galf as side chains of 3–4 units that are (1→5)-interlinked. The importance of the carbohydrate moiety of the F. oxysporum PGM was demonstrated. Periodate oxidation abolished much of the PGM antigenic activity. A strong decrease in reactivity was also observed with de-O-glycosylated PGM. In addition, de-O-glycosylated PGM was not able to inhibit F. oxysporum phagocytosis, suggesting that macrophages recognize and internalize F. oxysporum via PGM. F. oxysporum PGM triggered TNF-α release by macrophages. Chemical removal of O-linked oligosaccharides from PGM led

C The Author(s) 2018. Published by Oxford University Press on behalf of The International Society for Human and Animal Mycology. All rights reserved. For permissions, please e-mail: [email protected]

Downloaded from https://academic.oup.com/mmy/advance-article-abstract/doi/10.1093/mmy/myx167/4996173 by UFRJ user on 15 May 2018

1

2

Medical Mycology, 2018, Vol. 00, No. 00

to a significant increase of TNF-α cytokine levels, suggesting that their removal could exposure another PGM motifs able to induce a higher secretion of TNF-α levels. Interestingly, F. oxysporum conidia, intact and de-O-linked PGM were not able to induce IL-10 cytokine release. The difference in patient serum reativity using a PGM from F. oxysporum characterized in the present study as compared with a PGM from C. resinae, that presents the same epitopes recognized by serum from patients with aspergillosis, could be considered a potential diagnostic antigen and should be tested with more sera. Key words: Fusarium oxysporum, peptidogalactomannan, macrophage activation, potential diagnostic antigen.

Introduction Fusarium species are common hyaline soil saprophytes and plant pathogens. In humans, Fusarium species cause superficial infections in immnocompetent individuals, and severe and disseminated disease in immunocompromised hosts.1–4 Fusarium solani, F. oxysporum, and F. moniliforme are the most frequent species causing disease in humans.5 Structural studies on the cell wall glycoconjugates from Fusarium spp. are quite limited. In a sequence of studies initiated by Jikibara et al. (1992), uronic acid-containing glycoproteins were isolated from a Fusarium sp.6 Structurally different glycoprotein fractions were obtained from mycelia of F. oxysporum, following hot aqueous extraction.7 Recently, two diferent galactofuranose-containing antigens have been isolated from the cell wall and culture supernatants of Fusarium.8 Using monoclonal antibodies against these antigens and a galactomannan-specific antibody, hyphae from Fusarium and A. fumigatus could be differentiated by immunofluorescence and immunohistology techniques. As the antigen is released in the culture medium, it could be a target for a Fusarium-specific serological assay.8 The chemical characterization of fungal antigens is important to allow a rational interpretation of the crossreactivity between pathogenic species, thus helping selection of immunodominant components that can be of diagnostic use. The diagnosis of invasive mycoses is based on the growth of fungi in different biologic materials, as well as the demonstration of tissue invasion by fungi. In addition, detection of polysaccharide antigens in body fluids by immunologic assays has been advanced, and constitutes an important diagnostic tool in different clinical scenarios. The polysaccharides detected are galactomannans of Aspergillus spp.,9 and Histoplasma capsulatum,10 β(1→3)-glucan of various fungi, including Candida spp., Aspergillus spp., Fusarium spp., Acremonium spp., Pneumocystis jirovecii and others,11 and glucuronoxylomannan of Cryptococcus neofor-

mans.12 However, a major problem is the occurrence of false-positive results related in some cases to cross-reactivity with other fungi. This was confirmed for the commercial galactomannan test, Platelia EIA. The epitope, (1→5)-βD-galactofuranoside side chains of the Aspergillus galactomannan, detected by the EB-A2 monoclonal antibody employed in the test is not exclusively present in Aspergillus species. This epitope may be present in antigens of other fungi, including Penicillium, Paecilomyces, Trichothecium, Myceliophthora, Blastomyces dermatitidis, H. capsulatum, as well as Geotrichum capitatum, Trichosporon species, Acremonium species, Alternaria alternate and Fusarium species.13 Swanink et al. (1997) and Cummings et al. (2007) in contrast to the findings of Kappe and Schulze-Berge (1993), were unable to find cross-reactivity of F. oxysporum, probably due to differences in the antigen preparation among species or strains and some other factors.13–15 Tortorano et al. (2012) using 12 Fusarium isolates (F. oxysporum, F. verticillioides, F. solani, F. falciforme), reported the cross-reactivity of these Fusarium spp. exoantigens in the Aspergillus galactomannan enzyme-linked immunosorbent assay (ELISA) (Platelia Aspergillus assay).16 They conclude that a positive galactomannan test in an immunocompromised host may represent invasive aspergillosis or another fungal infection including Fusarium. More recently, Nucci et al. (2014) reported 15 out of 18 patients with invasive fusariosis who tested positive for the Platelia Aspergillus galactomannan antigen.17 The cross-reactivity of the Platelia Aspergillus galactomannan assay with Fusarium spp. may constitute a drawback for the specificity of this test, impacting the choice of antifungal therapy.17 The mechanism of Fusarium cross-reactivity is not yet elucidated. The management of invasive fusariosis is challenging because the outcome is poor and largely dependent on host defenses.18,19 In this study, a detailed investigation of the major glycoconjugate present on the F. oxysporum cell surface was carried out by a combination of chemical and spectroscopic methods, including one- and two-dimensional nuclear


de Oliveira et al.

magnetic resonance (1D and 2D NMR) spectroscopic analysis. Its structure was compared with other molecules found in the fungi A. fumigatus and C. resinae that synthesize antigenic galactomannans. We analyzed how the glycosylation of F. oxysporum glycoconjugate influences the recognition and uptake of F. oxysporum by macrophages as well as its role in the production of proinflammatory cytokines. In addition, the reactivity of the glycoconjugate was evaluated by ELISA using sera from patients with invasive fusariosis and invasive aspergillosis.

Methods Strain and culture conditions Fusarium oxysporum (IOC 4247) was kindly supplied by Maria Inês Sarquis from the Collection Culture of Institute Oswaldo Cruz, Rio de Janeiro, RJ, Brazil, and maintained on slants of Sabouraud medium. Cells were inoculated in Erlenmeyer flasks containing Sabouraud medium, which was incubated for 7 days at 25◦ C with shaking. The mycelium was obtained via paper filtration, washed with distilled water and stored at −20◦ C.

Extraction of Fusarium oxysporum mycelium and fractionation of glycoprotein with Cetavlon The crude glycoprotein was extracted from mycelium with 0.05 M phosphate buffer, pH 7.2 at 100◦ C for 2 h, the extract was dialyzed against distilled water, evaporated to a small volume and then fractionated by Cetavlon precipitation according to Haido et al. (1998).20 The mother liquor from the first Cetavlon precipitation was adjusted to pH 8.8 in the presence of borate and the resulting precipitate recovered by centrifugation to give a major glycoprotein fraction (peptidogalactomannan, PGM). The fraction was dialyzed against distilled water and freeze-dried.

Monosaccharide composition and methylation analysis of the glycoprotein Monosacharide composition in PGM was determined based on the chemical analysis and gas–liquid chromatography– mass spectrometry (GC-MS) analyses. The carbohydrate content was determined by the Dubois reaction, protein was determined by the Folin phenol reagent method and the presence of uronic acid was determined by the carbazole reaction.21–23 For the monosaccharide composition by GCMS, 2 mg of PGM were hydrolyzed with 5 M trifluoroacetic acid for 4 h at 100◦ C, reduced with borohydride, and the alditols acetylated with acetic anhydride: pyridine (1:1,v/v). The acetylated alditols dissolved in chloroform and ana-

3

lyzed in a GC-MS (GCMS-QP2010 Shimadzu, Japan) with a Restek column RTX-5MS, according to Kircher (1960).24 For methylation analysis the glycoprotein (5 mg) was subjected to two rounds of methylation as described by Ciucanu and Kerek (1984).25 The methylated glycoprotein was hydrolyzed, reduced with borohydride, acetylated, and analyzed on a GC-MS instrument, as described above.

NMR Proton and 13 carbon (1 H and 13 C) one- and twodimensional nuclear magnetic resonance (1D and 2D) spectra of PGM recorded using a 500 MHz NMR spectrometer (Bruker Biospin, Rheinstetten, Germany) with a triple resonance probe. Approximately 20 mg of each sample dissolved in 0.5 ml of 99.9% deuterium oxide (Cambridge Isotope Laboratory, Cambridge, MA, USA). All spectra recorded at 50◦ C with deuterated water exhibiting a peak due to exchange with residual H2 O (HOD) suppression by presaturation. For 1D 1 H NMR spectra, 64 scans were recorded using an inter-scan delay equals 1 s. For 2D 1 H–1 H COSY, phase sensitive 1H–1H TOCSY, 1 H–1 H NOESY, 1 H–13 C HSQC, and 1 H–13 C HMBC experiments, spectra were recorded using states time proportion phase incrementation (TPPI) for quadrature detection in the indirect dimension. Phase sensitive TOCSY spectra run with 4096 × 512 points with a spin lock field of 10 kHz and a mixing time of 60 ms. NOESY spectra run with 4046 × 512 points and multiple mixing times of 50, 100, 200, and 400 milliseconds were tested. Mixing time of 100 milliseconds was more appropriated for analysis of FCS. Two-dimensional 1 H–13 C Multiplicity-Edited HSQC spectra were recorded at 50◦ C with HOD suppression by presaturation, with 256 scans. The increment number setup was set to 64, and states-TPPI used for quadrature detection in the indirect dimension and run with 1024 × 512 points with globally optimized alternating phase rectangular pulses for decoupling. The 1 H–13 C HMBC spectra recorded with 1024 × 256 points, with a 60 ms delay for evolution of long-range couplings, and set with no decoupling during acquisition time. Chemical shifts displayed relative to external trimethylsilylpropionic acid at 0 ppm for 1 H and relative to methanol for 13 C. The data were processed using TopSpin3.1 (Bruker Biospin, Billerica, MA, USA).26

Partial hydrolysis of the glycoprotein from F. oxysporum Fusarium oxysporum PGM (2 mg) was treated with 0.1 M HCl and heated at 100◦ C for 20 min. The degraded glycoprotein was recovered on dialysis against distilled water and freeze-drying of the retained solution.27


4

Periodate oxidation of the glycoprotein Fusarium oxysporum PGM (4 mg) was treated with increasing concentrations (10–100 mM) of sodium m-periodate and the reaction mixture was kept for 18 h at 4◦ C in the dark. The reaction was stopped by adding an equimolar amount of glycerol and then, after 15 min, 100 mM sodium borohydride. After incubation for 2 h at 4◦ C, the reaction mixture was dialyzed against distilled water at 4◦ C for 24 h.20

Beta-elimination of the glycoprotein Fusarium oxysporum (5 mg) was chemically de-Oglycosylated by mild reductive alkaline treatment under reducing conditions (0.1 M NaOH, 0.5 M NaBH4 , 25◦ C, 24 h). The de-O-glycosylated PGM was purified by gel permeation chromatography in a Bio-Gel P-2 column (2 × 140 cm) being recovered in the void volume.28

Rabbit immune sera White male rabbits were inoculated with freeze-dried whole cells of F. oxysporum (2 mg/ml dry weight) emulsified in an equal volume of complete Freund´s adjuvant. In sum, 1 ml of emulsion was injected intradermally at weekly intervals of 3 weeks.20 Then, throughout a 1-week period the same concentration was used in three intravenous injections at 2day intervals. The hyperimmune serum obtained was used in ELISA experiments.

Human sera Human sera were from the Mycology Laboratory of the University Hospital, Universidade Federal do Rio de Janeiro. These sera were from patients with aspergillosis (03) and fusariosis (05). The number of healthy control sera was 08.

ELISA Solutions of 5 μg/ml of F. oxysporum PGM (dry weight), intact and de-O-glycosylated were added onto wells of flat-bottomed polyvinyl microtiter plates (Falcon-Becton & Dickinson, Franklin Lakes, NJ, USA), which were maintained for 1 h at 37◦ C and overnight at 4◦ C. After washing the plates with 0.05% phosphate buffered saline-Tween 20, the nonspecific sites were blocked by addition of 5% skimmed milk in 0.1% phosphate buffered saline-Tween 20. Human and rabbit antisera (1/400) in blocking buffer (100 μl) were added to the wells, and antibody binding was measured using, respectively, goat anti-human im-


munoglobulin G (IgG) and goat anti-rabbit IgG antibodies conjugate to horseradish peroxidase (Sigma). The chromogen used was O-phenylenediamine added together with H2 O2 . Periodate treated and de-O-glycosylated glycoprotein were tested by ELISA as described above.

Phagocytic assay Elicited peritoneal macrophages were obtained according to Xisto et al. (2015).29 Elicited macrophages (5 × 105 cells/well) were cultured over round glass coverslips (13 mm) in 24-well flat bottom microtest plates. Adherent monolayers of primary cells were challenged with 500 μl of live conidia suspensions containing 2.5 × 106 cells/well. After incubation at 37◦ C in 5% CO2 for 2 h in RPMI 1640 medium, the cells were rinsed with Roswell Park Memorial Institute (RPMI) medium for removal of noninternalized conidia. The preparations were fixed in Bouin’s fixative and stained with Giemsa. The influence of intact and de-Oglycosylated PGM on conidia phagocytosis was evaluated by adding 50 μg/ml of glycoprotein 30 min before the addition of conidia. To determine the phagocytic ´ındexes (PIs), 200 cells were counted, and the percentage of cells that ingested at least one particle was multiplied by the mean number of internalized particles.30

Macrophage viability assay—Neutral red dye-uptake method Solutions of 7.8 to 500 μg/ml of intact and de-Oglycosylated glycoprotein (dry weight) were added to macrophages plated in 96-well plates. After 24 h, the cytotoxic effect on macrophages was analyzed by the neutral red technique.29

Macrophage effector function—Cytokine assay The cytokine assay was conducted under the same conditions described above for the phagocytosis assay. Adherent cells were stimulated for 18 h in RPMI medium, with the intact and the de-O-glycosylated glycoprotein, heatkilled conidia (ratio 5:1) or 10 ng/well of lipopolysaccharide (LPS) from Escherichia coli O111:B4 (Sigma Aldrich, St. Louis, MO, USA). After this period, the supernatant was recovered for tumor necrosis factor α (TNF-α) and interleukin 10 (IL-10) determination by ELISA according to the manufacturer´s instructions. Polymixin B (10 μg/ml) was added 5 min before the addition of the stimulus, to rule out the possibility that the stimulating activity was due to contaminating LPS. After incubation for 18 h, supernatants were harvested, centrifuged at 12 000 rpm for 10 min to remove cell debris, and stored in cryogenic vials at −80◦ C. In


de Oliveira et al.

the supernatants obtained the concentration of TNF-α and IL-10 was measured by ELISA (BD OptEIA, Mouse TNFα, and IL-10 ELISA Set) according to the guidelines of the manufacturer.

Ethics statement ´ The study was approved by Comite de Etica em Pesquisa (CEP), Hospital Universitario Clementino Fraga Filho ´ (HUCFF), of the Federal University of Rio de Janeiro, ´ Brazil, Process no,. 021/07 and by Comite de Etica no Uso de Animais (CEUA), of the Federal University of Rio de Janeiro State (Universidade Federal do Estado do Rio de Janeiro – UNIRIO), Brazil, Process no. 004/2014.

Statistical analysis Statistical analyses were performed using GraphPad Prism version 5.00 for Windows (GraphPad Software, San Diego, CA, USA). One-way ANOVA was used to compare differences between groups, and individual comparisons of groups were made using the Bonferoni test (Bonferoni posttest) (P < .05).

Results The hot aqueous extract of F. oxysporum mycelium was isolated and treated with Cetavlon plus sodium tetraborate at pH 8.8. A precipitate of PGM was formed which contained neutral carbohydrate (85%) and protein (12%). The monosaccharide composition was determined by GC-MS analysis. We detected mannose, galactose and glucose in a 2:1:1 molar ratio and traces of uronic acid, detected by colorimetric method.

Methylation analysis For the determination of the linkage position the partially O-methylated alditol acetates were analyzed by GC-MS. We observed principally mannopyranosyl and glucopyranosyl nonreducing end units, with 2-O- substituted D-Glcp, 2-O- and 6-O- substituted D-Manp and 2,6-di-O- substituted Galf units.

NMR spectroscopy Analysis of the 1D NMR spectrum of F. oxysporum PGM demonstrated a huge number of superimposed signals that could be elucidated using 2D NMR analysis. Using 2D NMR techniques such as 1 H-1 H COSY, TOCSY, and 2D 1 H -13 C (HSQC and HMBC) together with literature data,31 the PGM structure could be elucidated.

5

In the anomeric region of the HSQC spectrum, eight signals appeared at 5.12/106.71, 5.01/98.55, 5.01/98.6, 4.77/101.44, 4.71/101.3, 4.55/100.41, 4.96/102.71 and 4.86/99.89 (Fig. 1, Table 1). Typical shifts corresponding to β-Galf units at 5.12/106.71 ppm (A1 unit). Based in the methylation analysis (data not shown), COSY, TOCSY and HSQC spectra, anomeric signals at 5.01/98.55 (B1 unit) correspond to α-glucopyranose nonreducing end unit (2,3,4,6-Me4-Glc) and a superimposed shifts at 5.01/98.6 (C1 unit), correspond to a 2-O-substituted α-D-Glcp units (3,4,6-Me3-Glc). Minor signals were detected for units D and E, presenting similar chemical shifts also observed by Chen et al. (2015) from an extracelular polysaccharide isolated from mangrove-associate F. oxysporum and that correspond to β-D-Manp (1→ and →2) - β-D-Manp (1→ respectively (2,3,4,6-Me4-Man and 3,4,6-Me3-Man). Unit G showed an anomeric signal at 103.1 ppm that corresponds to α-D-Manp - nonreducing end units30 and unit H presented a signal at δ 99.89 corresponding to α-D-Manp (1→6)- substituted units (2,3,4-Me3-Man).31 The HSQC spectrum differs from a PGM previously isolated from C. resinae (Fig. 2), suggesting the presence of a different PGM in F. oxysporum. The sequence of glycosidic linkages presented in the F. oxysporum PGM could be determined based on the TOCSY phase sensitive and HMBC spectra. In the TOCSY spectrum (Fig. 3A), it could be observed in the antiphase ROE’s, that suggest the glycosidic linkages. The positions of the glycosidic linkages could be confirmed by the HMBC spectrum (Fig. 3B). In this spectrum, connecting points between A1 (106.71) and A6 H (3.59–3.84) could be detected and represent the PGM main chain. This chain is substituted at 2-position by B, C, D, E, F, and G (B1C(98.55) and A2H (4.12) , C1C(98.6) and A2 H(4.12) , D1 C(101.55) - A2 H (4.12) , E1 C (106.77) and A2 H (4.12) , F1 C (100.47) and between G1 C (102.71) and A2 H (4.12) . (B and C are the major units). Also detected was contact point between H1 C(99.89) and H6 H (3.71–3.91) suggesting the presence of (→ 6)- α-Manp (1→chain. These results are in agreement with methylation analysis and the suggested structure of PGM consists of a backbone of (1→6)-linked β-D-galactofuranose residues with several branches at C-2 containing α-D-Glcp (1→, β-D-Manp (1→2)- β-D-Manp- α-D-Glcp (1→ and β-DManp- α-D-Glcp (1→. Structural fragments present in the PGM of F. oxysporum are shown in Figure 4 (a, b, c, d, and e).

The role of carbohydrate epitopes in antigenic reactivity of purified PGM To evaluate the role of the carbohydrate epitopes of the PGM in the ELISA reaction, rabbit anti-serum was


6


Figure 1. HSQC (1 H-13 C) spectrum of the PGM from F. oxysporum, showing the signals of CH carbon/proton (phase signals) and those from CH2 (antiphase). The signals and corresponding units were identified in Table 1. This Figure is reproduced in color in the online version of Medical Mycology.

tested against PGM submitted to periodate oxidation, and β-elimination treatment (removal of O-linked oligosaccharides). The antigenic epitopes were, in large part, carbohydrate in nature. Periodate oxidation abolished much of the PGM antigenic activity (Fig. 5A). A strong decrease in reactivity was also observed with de-O-glycosylated PGM (lacking O-linked oligosaccharide chains) (Fig. 5B).

Involvement of PGM on phagocytosis of F. oxysporum by macrophages To evaluate whether PGM and its O-linked oligosaccharides are involved in F. oxysporum uptake, macrophages were pretreated with 50 μg/ml intact and de-O-glycosylated PGM for 30 min before interaction with F. oxysporum conidia for 2 h. The conidia to macrophage ratio was adjusted to 5:1. Pretreatment of the macrophages with intact PGM led to the inhibition of conidia phagocytosis, whereas de-O-glycosylated PGM did not show any inhibition on macrophage conidia internalization (Fig. 6). These results reveal that macrophages recognize and internalize F. oxysporum via PGM, and demonstrate that O-glycosylation

Table 1. Assignment of HSQC (1 H/13 C) chemical shift values (anomeric region) of F. oxysporum PGM. Unit

Structure

H1/C1

A1 B1 C1 D1 E1 F1 G1 H1

→2,6)-β-D-Galf (1→ α-D-Glcp (1→ →2)-α-D-Glcp (1→ β-D-Manp (1→ →2)-β-D-Manp (1→ α-D-GlcpA (1→ α-D-Manp (1→ →6)-α-D-Manp (1→

5.12/106.71 5.01/98.55 5.01/98.6 4.77/101.44 4.71/101.3 4.55/100.41 4.96/102.71 4.86/99.89

plays a role in this process. These results are in agreement with the previous one29 using a peptidorhamnomannan from S. prolificans and its de-O-glycosylated derivative.

Cytotoxic assay of F. oxysporum PGM The cytotoxicity of soluble PGM was assessed by the neutral red dye-uptake method. At PGM and de-O-glycosylated PGM concentrations, ranging from 7.8 to 500 μg/ml the macrophages viability was more than 90%.


de Oliveira et al.

7

Table 2. Assignment of HSQC (1 H/13 C) chemical shift values (anomeric region) of C. resinae and F. oxysporum. C. resinae units

Structure

H1/C1

A B C D E F

→5)-β-D-Galf (1→ β-D-Galf (1→ →6)-α-D-Manp (1→ →2)-α-D-Manp (1→ α-D-Manp (1→ →2,6)-α-D-Manp(1→

5.12/107.34 5.01/106.6 4.95/99.61 5.14/101.0 4.96/102.62 4.94/97.72

F. oxysporum units

Structure

H1/C1

A B C D E F G H

→2,6)-β-D-Galf (1→ α-D-Glcp (1→ →2)-α-D-Glcp (1→ β-D-Manp (1→ →2)-β-D-Manp (1→ α-D-GlcpA (1→ α-D-Manp (1→ →6)-α-D-Manp (1→

5.12/106.71 5.01/98.55 5.01/98.6 4.77/101.44 4.71/101.3 4.55/100.41 4.96/102.71 4.86/99.89

Pro-inflammatory cytokine release induced by F. oxysporum PGM The role PGM plays in F. oxysporum conidia induction of TNF-α, a distinct proinflammatory cytokine, and of the anti-inflammatory IL-10, by macrophages was examined (Fig. 7). The TNF-α cytokine was produced by the macrophages stimulated with conidia and with intact PGM (Fig. 7A). When de-O-glycosylated PGM was tested, a significant increase in cytokine levels was observed for TNF-α (Fig. 7A). However, conidia, intact or de-O-glycosylated PGM were not able to stimulate IL-10 cytokine secretion (Fig. 7B).

Figure 2. HSQC (1 H-13 C) anomeric region of PGM from C. resinae and F. oxysporum. The signals and corresponding units was identified in Table 2. This Figure is reproduced in color in the online version of Medical Mycology.

Discussion Reactivity of purified PGM from Fusarium and Cladosporium with serum from patients with invasive fusariosis and invasive aspergillosis: Effect of Galf (1→5) Galf removal on serum reactivity PGM from F. oxysporum and C. resinae were tested by ELISA with serum from patients with invasive fusariosis and invasive aspergillosis (Fig. 8A). Treatment with dilute acid, which removed labile galactofuranosyl sidechain residues, decreased the antigenicity of Cladosporium PGM (Fig. 8B). In contrast, no significant difference was observed in Fusarium PGM showing that this molecule does not contain Galf (1→5) Galf moieties (Fig. 8B) as confirmed by structural analysis.

β-Galactofuranose containing structures were characterized by our group in A. fumigatus, A. versicolor, A. flavus, and C. resinae.20,27,28,32,33 Typically, Galf has been described to occur in the galactomannan as side chains with repeating β- (1→5) -Galf units that are 2, 6 or 3, 6 linked to the mannose backbone. We now report the structural characterization of a PGM from F. oxysporum by a combination of chemical and spectroscopic methods, including 1D and 2D NMR presenting a diferent structure as the one previously described in Aspergillus species. Our data obtained by NMR and methylation analysis suggest that the F. oxysporum PGM structure consists of a main chain containing (1→6)linked β-D-galactofuranose residues with side chains


8


Figure 3. NMR analysis linkage position of PGM. (A) Phase-sensitive 1 H-1 H TOCSY spectrum. Signals of the spin systems are on in-phase, whereas signals from the neighboring residues are in anti-phase. (B) 13 C-1 H HMBC spectrum showing the contact points between saccharide units of PGM. This Figure is reproduced in color in the online version of Medical Mycology.

Figure 4. Main structures of PGM. The galactomannan component of Fusarium PGM consists of a main- chain containing (1→6)-linked- β-D-Galf residues (c) with several branches at C-2 containing α-D-Glcp (1→ (a), α-D-Manp(1→ (e) and short chains containing β-D-Manp (1→2)- Manp (b) and β-D-Manp (1→2)- Manp-(1→2)-α-D-Glcp (d).

containing (1→2)-linked α-D-Glcp, (1→2)-linked -β-DManp (1→2) and β-D-Manp terminal nonreducing end units. Such a complex galactomannan is very similar to an extracellular galactomannan with antioxidant activity from the mangrove-associated F. oxysporum31 and the Nand O-linked glycan fractions of F. oxysporum cell wall

proteins.34 It differs from PGMs isolated from A. fumigatus20 and C. resinae27 that present a main chain containing (1→6)-linked α-D-Manp residues substituted at O-2 by side chains containing (1→2)-linked α-D-Manp residues. β –D-Galf residues were present as side chains of 3–4 units that are (1→5)-interlinked, as well as the galactomannan


de Oliveira et al.

9

Figure 5. Effect of different treatments in antigenic reactivity of F. oxysporum PGM. Untreated, de-O-glycosylated and periodate treated PGM (5 μg/ml) were probed with rabbit anti-serum in ELISA assay. Rabbit pre-immune serum was used as negative control. (A) Periodate oxidation (50 mM/100 mM) abolished much of the PGM antigenic activity (B) A strong decrease in reactivity was also observed with de-O- glycosylated PGM. Values represent the mean ± SD of three independent experiments performed in triplicate. Asterisks denote values statistically different from control (∗ P < .05; ∗∗∗ P < .001; NS, no significance).

Figure 6. Inhibition assay of the phagocytosis between F. oxysporum conidia and mouse peritoneal macrophages by intact and de-Oglycosylated PGM from F. oxysporum. The macrophages were treated or not (control) with intact and de-O-glycosylated PGM (50 μg/ml), before the interaction with conidial cells (ratio 5:1) for 2 h. The phagocytic index values represent the mean ± SD of three independent experiments performed in triplicate. Asterisks denote values statistically different from control (∗ P < .05; ∗∗ P < .01).

isolated from C. werneckii, the causative agent of tinea nigra having a single terminal β-galactofuranosyl-(1→6)units35 and from (1→6)-linked linear β-galactofuranosyl polymers linked to a small amount of mannan, isolated from Malassezia furfur and M. pachydermatis.36 The importance of the carbohydrate moiety of F. oxysporum PGM in serum recognition has been demonstrated in this work. Treatment of PGM with sodium metaperiodate and release of O-linked oligosaccharides by reductive beta elimination reaction strongly decreased the antibody binding capacity. These results showed that the carbohydrate portion of PGM is predominantly responsible for

the antigenic activity and O-linked oligosaccharides are important determinants of reactivity as already described in peptidopolysaccharides from A. fumigatus,20,28 Pseudallescheria boydii,37 Sporothrix schencki,38 and C. resinae.27 Using reverse-genetic and biochemical approaches, Komach et al. (2013) characterized a galactofuranosyl transferase-encoding gene (gfsA) with the ability to synthesize the Galf antigen of O-glycans in A. nidulans and A. fumigatus.39 Recently, Katafuchi et al. (2017) using LC/MS, 1 H-NMR and methylation analysis of the enzymatic products showed that AfGfsA has the ability to transfer the Galf to the C-5 position of the β-Galf residue via β-linkage.40 An interesting observation previously reported by Leitao ˜ et al. (2003) was that O-linked oligosaccharides contained some structural features were also present in the polysaccharides (supposedly N-linked glycans), or for instance, chains terminated by (1→5)-linked β-Galf residues and having one of them substituting O-6 of mannosyl residues.28 This suggests that the same glycosyl transferases that modify N-glycans may also modify N-glycans in A. fumigatus, as occurs in Saccharomyces cerevisiae.41 To detect the occurrence of O-glycosylation in a growing number of medically important fungi may be an important step toward the better understanding of their physiology and may be useful for providing potential targets for drug development. The findings of Komach et al. (2013) and Katafuchi et al. (2017) about the GfsA suggest that GfsA activity maybe a potential target for antifungal treatment.39,40 We also demonstrated that O-linked oligosaccharides are the key determinants for the phagocytosis of conidia by murine macrophages and induction of the inflammatory response. Intact PGM from F. oxysporum inhibits the phagocytosis of conidia. However, de-O-glycosylated


10


Figure 7. Cytokine release induced by conidia, PGM and PGM de-O-glycosylated from F. oxysporum. (A) Both F. oxysporum conidia and PGM (50 μg/ml) were able to stimulate the production of TNF-α by macrophages. However, the TNF-α levels were significantly increased when de-Oglycosylated PGM was tested (B). IL-10 release by macrophages when stimulated with conidia, PGM or de-O-glycosylated PGM were not altered. Values are means of three independent experiments. Asterisks denote values statistically different from control (MO) (∗∗ P < .01; ∗∗∗ P < .001; NS, no significance).

PGM (lacking O-linked chains) was not able to inhibit F. oxysporum phagocytosis, suggesting that F. oxysporum O-linked chains are the PGM moiety recognized by phagocytic receptors. These results are in agreement with the previous one29 showing that O-linked chains from peptidorhamnomannan are determinants for uptake of S. prolificans by macrophages. The TNF-α cytokine was produced by the macrophages stimulated with PGM. When de-O-glycosylated PGM was tested, a significant increase in TNF-α cytokine levels was observed as compared with the intact PGM. Our results suggest that the removal of O-linked oligosaccharides from PGM could exposure another PGM motifs able to stimulate a higher TNF-α level compared with the intact PGM molecule. Interestingly, F. oxysporum conidia, intact and de-O-glycosylated PGM were not able to induce IL-10 cytokine release. Our results imply that the differential recognition of PGM motifs by macrophages can lead to the activation of different cellular events. Previous work has been demonstrated that conidia and N-linked rhamnomannans from P. boydii induced the production of TNF-α and IL-10 cytokines through TLR4 signaling with MAPK phosphorylation.42 S. prolificans PRM, on the other hand, is able to stimulate TNF-α but not IL-10.29 In Candida albicans, O-linked mannans induce innate immune activation associated with pro-inflammatory cytokine release via TLR4.43 In addition to TLR4, Dectin-2 is a receptor involved in the recognition of C. albicans α-mannans.44,45 A. fumigatus galactomannan can be recognized by dectin 2 leading to important antifungal response and the induction of cytokines. Furthermore, dendritic-cell-specific ICAM3-grabbing non-

integrin (DC-SIGN), has been hypothesized to recognize galactomannan.46 To confirm our structural analysis that shows the absence of β-(1→5)-linked galactofuranosyl oligosaccharide side chains in F. oxysporum PGM, previously reported in several fungal galactomannans,20,27,28 treatment of PGM from C. resinae27 and F. oxysporum with HCl 0.1 M, which removed labile galactofuranosyl side-chain residues were done and the partially hydrolyzed PGM were analyzed by NMR (data not shown) and their reactivity were tested on ELISA with serum from patients with invasive fusariosis and invasive aspergillosis. The immunodominant epitopes previously identified in A. fumigatus28 and C. resinae27 as tetra- and hexasaccharides, which contain a β-Galf-(1→5) - β-Galf- terminal groups, and absent in Fusarium PGM, were removed, and a decrease in reactivity was observed (Fig. 8B). In conclusion, our results showed that galactofuranosyl residues typically occur in the PGM as side chains with repeating β-Galf-(1→5) units linked to a mannose backbone consisting of (1→6)-linked α-D-mannopyranosyl residues.20,33 However, variation in this structure has been reported in this study, differing both in the mannan core and in the side chains. PGM from Fusarium have consecutive β (1→6)-linked Galf residues, O-2 substituted by side chains of β-D-Manp and α-D-Glcp. A cross-reactivity was observed between PGM from F. oxysporum and PGM from C. resinae (Fig. 8A), suggesting the presence of another structural feature of these PGMs recognized by patients sera. Cross-reactivity between PGM from C. resinae and mannan from Candida parapsilosis was also observed and the presence of side chains of (1→2)-linked α-D-mannopyranosyl


de Oliveira et al.

11

and specific, and appears in the serum before the clinical manifestations of invasive fusariosis may have an impact in decreasing the mortality of this devastating disease. The PGM characterized in the present study may be a candidate, and should be tested with more sera.

Acknowledgments This work was supported by the Conselho Nacional de Desenvolvimento Cient´ıfico e Tecnologico (CNPq), Fundaçao ´ ˜ de Amparo a` Pesquisa do Estado do Rio de Janeiro (FAPERJ), Coordenaçao ˜ de Aperfeiçoamento de Pessoal de N´ıvel Superior (CAPES-PROEX), and Universidade Federal do Rio de Janeiro (UFRJ).

Declaration of Interest The authors report no conflicts of interest. The authors alone are responsible for the content and the writing of the paper.

References 1.

2.

3. 4. Figure 8. Reactivity of purified PGM from Fusarium and Cladosporium and the effect of Galf (1→5)-side chains removal. (A) PGM from F. oxysporum and C. resinae (5 μg/ml) were tested on ELISA with serum (1:200) from patients with invasive fusariosis and invasive aspergillosis. (B) Treatment with dilute acid (HCl 0.1 M) decreased the antigenicity of Cladosporium PGM. In contrast, no significant difference was observed in Fusarium PGM.

5.

units in the C. resinae PGM and in C. parapsilosis could be involved in this reactivity.27 Wiedemann et al. (2016) detected a Galf-containing antigen of unknown structure in supernatants of several Fusarium species using a novel monoclonal antibody.8 This antigen is also present in minor amounts in hyphae of A. fumigatus. The presence of distinct galactofuranose antigens in the cell wall as well as in culture supernatants is proposed as a strategy to differentiate Fusarium from Aspergillus species. The difference in patient serum reativity using a PGM from F. oxysporum as compared with a PGM from C. resinae, which presents the same epitopes recognized by serum from patients with aspergillosis, could be considered. The outcome of invasive aspergillosis has improved in the last decade, and one of the reasons is the early diagnosis of the disease, achieved after the introduction of serum galactomannan testing.47 With this regard, a test that is sensitive

8.

6.

7.

9.

10.

11. 12.

13.

14.

Anaissie E, Nelson P, Beremand M, Kontoyiannis D, Rinaldi M. FusariumCaused Hyalohyphomycosis: An Overview. In: Borgers M, Hay R, Rinaldi MG, eds. Current Topics in Medical Mycology. New York: Springer; 1992: 231–249. Guarro J. Lessons from animal studies for the treatment of invasive human infections due to uncommon fungi. J Antimicrob Chemother. 2011; 66: 1447–1466. Guarro J, Gené J. Opportunistic fusarial infections in humans. Eur J Clin Microbiol Infect Dis. 1995; 14: 741–754. Nucci M, Anaissie E. Fusarium infections in immunocompromised patients. Clin Microbiol Rev. 2007; 20: 695–704. Nucci M, Marr KA, Vehreschild MJGT et al. Improvement in the outcome of invasive fusariosis in the last decade. Clin Microbiol Infect. 2014; 20: 580–585. Jikibara T, Takegawa K, Iwahara S. Studies on the uronic acid-containing glycoproteins of Fusarium sp. M7-1: I. Isolation and some properties of the glycoproteins. J Biochem. 1992; 111: 225–229. da Silva RB, Ribeiro SA, Sassaki GL, Gorin PAJ, Barreto-Bergter E. Uronic acid-containing glycopeptides from Fusarium oxysporum: Possible significance as chemotypes. Carbohydr Polym. 2005; 60: 449–455. Wiedemann A, Kakoschke TK, Speth C et al. Distinct galactofuranose antigens in the cell wall and culture supernatants as a means to differentiate Fusarium from Aspergillus species. International J Med Microbiol. 2016; 306: 381–390. Eigl S, Hoenigl M, Spiess B et al. Galactomannan testing and Aspergillus PCR in same-day bronchoalveolar lavage and blood samples for diagnosis of invasive aspergillosis. Med Mycol. 2017; 55: 528–534. Theel ES, Jespersen DJ, Harring J, Mandrekar J, Binnicker MJ. Evaluation of an enzyme immunoassay for detection of Histoplasma capsulatum antigen from urine specimens. J Clin Microbiol. 2013; 51: 3555–3559. Schuetz AN. Invasive fungal infections: biomarkers and molecular approaches to diagnosis. Clin Lab Med. 2013; 33: 505–525. Percival A, Thorkildson P, Kozel TR. Monoclonal antibodies specific for immunorecessive epitopes of glucuronoxylomannan, the major capsular polysaccharide of Cryptococcus neoformans, reduce serotype bias in an immunoassay for cryptococcal antigen. Clin Vac Immunol. 2011; 18: 1292–1296. Kappe R, Schulze-Berge A. New cause for false-positive results with the Pastorex Aspergillus antigen latex agglutination test. J Clin Microbiol. 1993; 31: 2489–2490. Cummings JR, Jamison GR, Boudreaux JW, Howles MJ, Walsh TJ, Hayden RT. Cross-reactivity of non-Aspergillus fungal species in the


12

15.

16.

17.

18.

19. 20.

21. 22.

23. 24. 25. 26.

27.

28.

29.

30.

31.


Aspergillus galactomannan enzyme immunoassay. Diagn Microbiol Infect Dis. 2007; 59: 113–115. Swanink CM, Meis JF, Rijs AJ, Donnelly JP, Verweij PE. Specificity of a sandwich enzyme-linked immunosorbent assay for detecting Aspergillus galactomannan. J Clin Microbiol. 1997; 35: 257–260. Tortorano AM, Esposto MC, Prigitano A et al. Cross-reactivity of Fusarium spp. in the Aspergillus galactomannan enzyme-linked immunosorbent assay. J Clin Microbiol. 2012; 50: 1051–1053. Nucci M, Carlesse F, Cappellano P et al. Earlier diagnosis of invasive fusariosis with Aspergillus serum galactomannan testing. PLoS ONE. 2014; 9: e87784. Horn DL, Freifeld AG, Schuster MG, Azie NE, Franks B, Kauffman CA. Treatment and outcomes of invasive fusariosis: review of 65 R registry. Mycoses. 2014; 57: 652– cases from the PATH Alliance 658. Muhammed M, Coleman JJ, Carneiro HA, Mylonakis E. The challenge of managing fusariosis. Virulence. 2011; 2: 91–96. Haido RMT, Silva M, Ejzemberg R et al. Analysis of peptidogalactomannans from the mycelial surface of Aspergillus fumigatus. Med Mycol. 1998; 36: 313–321. Bitter T, Muir HM. A modified uronic acid carbazole reaction. Anal Biochem. 1962; 4: 330–334. DuBois M, Gilles KA, Hamilton JK, Rebers PA, Smith F. Colorimetric method for determination of sugars and related substances. Anal Chem. 1956; 28: 350–356. Lowry OH, Rosebrough NJ, Farr AL, Randall RJ. Protein measurement with the folin phenol reagent. J Biol Chem. 1951; 193: 265–275. Kircher HW. Gas-liquid partition chromatography of methylated sugars. Anal Chem. 1960; 32: 1103–1106. Ciucanu I, Kerek F. A simple and rapid method for the permethylation of carbohydrates. Carbohydr Res. 1984; 131: 209–217. Santos GRC, Glauser BF, Parreiras LA, Vilanova E, Mourao ˜ PAS. Distinct structures of the α-fucose branches in fucosylated chondroitin sulfates do not affect their anticoagulant activity. Glycobiology. 2015; 25: 1043– 1052. Calixto R, Mattos B, Bittencourt V et al. β-Galactofuranose-containing structures present in the cell wall of the saprophytic fungus Cladosporium (Hormoconis) resinae. Res Microbiol. 2010; 161: 720–728. Leitao ˜ EA, Bittencourt VCB, Haido RMT et al. β-Galactofuranosecontaining O-linked oligosaccharides present in the cell wall peptidogalactomannan of Aspergillus fumigatus contain immunodominant epitopes. Glycobiology. 2003; 13: 681–692. Xisto MIDS, Bittencourt VCB, Liporagi-Lopes LC et al. O-glycosylation in cell wall proteins in Scedosporium prolificans is critical for phagocytosis and inflammatory cytokines production by macrophages. PLoS One. 2015; 10: e0123189. Bittencourt VCB, Figueiredo RT, da Silva RB et al. An α-glucan of Pseudallescheria boydii is involved in fungal phagocytosis and Toll-like receptor activation. J Biol Chem. 2006; 281: 22614–22623. Chen Y-L, Mao W-J, Tao H-W et al. Preparation and characterization of a novel extracellular polysaccharide with antioxidant activity, from the mangrove-associated fungus Fusarium oxysporum. Mar Biotechnol. 2015; 17: 219–228.

32. da Silva Bahia MCF, Haido RMT, Figueiredo MHG et al. Humoral immune response in aspergillosis: an immunodominant glycoprotein of 35 kDa from Aspergillus flavus. Curr Microbiol. 2003; 47: 163–168. 33. da Silva Bahia MCF, Vieira RP, Mulloy B, Hartmann R, Bergter EB. The structures of polysaccharides and glycolipids of Aspergillus fumigatus grown in the presence of human serum. Mycopathologia. 1997; 137: 17–25. 34. Lopez-Fernandez L, Roncero MIG, Prieto A, Ruiz-Roldan C. Comparative proteomic analyses reveal that Gnt2-mediated N-glycosylation affects cell wall glycans and protein content in Fusarium oxysporum. J Proteom. 2015; 128: 189–202. 35. Lloyd KO. Isolation, characterization, and partial structure of peptido galactomannans from the yeast form of Cladosporium werneckii. Biochemistry. 1970; 9: 3446–3453. 36. Shibata N, Saitoh T, Tadokoro Y, Okawa Y. The cell wall galactomannan antigen from Malassezia furfur and Malassezia pachydermatis contains β-1,6-linked linear galactofuranosyl residues and its detection has diagnostic potential. Microbiology. 2009; 155: 3420–3429. 37. Pinto MR, Gorin PAJ, Wait R, Mulloy B, Barreto-Bergter E. Structures of the O-linked oligosaccharides of a complex glycoconjugate from Pseudallescheria boydii. Glycobiology. 2005; 15: 895–904. 38. Bernardes-Engemann AR, Loureiro y, Penha CV, Benvenuto F et al. A comparative serological study of the SsCBF antigenic fraction isolated from three Sporothrix schenckii strains. Med Mycol. 2009; 47: 874–878. 39. Komachi Y, Hatakeyama S, Motomatsu H et al. gfsA encodes a novel galactofuranosy ltransferase involved in biosynthesis of galactofuranose antigen of O-glycan in Aspergillus nidulans and A. fumigatus. Mol Microbiol. 2013; 90: 1054–1073. 40. Katafuchi Y, Li Q, Tanaka Y et al. GfsA is a β1,5galactofuranosyltransferase involved in the biosynthesis of the galactofuran side chain of fungal-type galactomannan in Aspergillus fumigatus. Glycobiology. 2017; 27: 568–581. 41. Lussier M, Sdicu A-M, Bussey H. The KTR and MNN1 mannosyltransferase families of Saccharomyces cerevisiae. Biochim Biophys Acta. 1999; 1426: 323–334. 42. Figueiredo RT, Fernandez PL, Dutra FF et al. TLR4 recognizes Pseudallescheria boydii conidia and purified rhamnomannans. J Biol Chem. 2010; 285: 40714–40723. 43. Netea MG, Gow NAR, Munro CA et al. Immune sensing of Candida albicans requires cooperative recognition of mannans and glucans by lectin and Toll-like receptors. J Clin Invest. 2006; 116: 1642–1650. 44. Saijo S, Ikeda S, Yamabe K et al. Dectin-2 recognition of α-mannans and induction of Th17 cell differentiation is essential for host defense against Candida albicans. Immunity. 2010; 32: 681–691. 45. Zhu Z, Wang H, Shang Q, Jiang Y, Cao Y, Chai Y. Time course analysis of Candida albicans metabolites during biofilm development. J Proteome Res. 2013; 12: 2375–2385. 46. van de Veerdonk FL, Gresnigt MS, Romani L, Netea MG, Latgé J-P. Aspergillus fumigatus morphology and dynamic host interactions. Nat Rev Microbiol. 2017; 15: 661. 47. Nucci M, Nouér SA, Cappone D, Anaissie E. Early diagnosis of invasive pulmonary aspergillosis in hematologic patients: an opportunity to improve the outcome. Haematologica. 2013; 98: 1657–1660.
