Annotation of Carbohydrate Active enZymes (CAZymes)
To compare the sugar-cleaving capabilities of the fungi selected, we have listed the number of representatives of each of the glycoside hydrolases (GH) and polysaccharide lyases (PL) families (defined in the carbohydrate active enzyme database (CAZy database) http://www.cazy.org;  and then performed a double clustering based on Bray-Curtis distances (i) between organisms according to their family distribution and (ii) between families according on their distribution pattern in the different genomes. Distances were computed using the multivariate analysis programme GINKGO . The distance matrixes were computed using the Species Profiles Distance model and analysed by a hierarchical agglomerative clustering method.
Fungal strains and enzymatic preparations
The fungal strains used in this study were obtained from the CBS-KNAW (Centraalbureau voor Schimmelcultures, Utrecht, The Netherlands) and the FGSC (Fungal Genetics Stock Center, Kansas City, USA) fungal collections (Table 1). They were streaked and verified by ITS (Internal Transcribed Spacer) sequencing and maintained in the culture collection entitled "Centre International de Ressources Microbiennes", which is dedicated to filamentous fungi of biotechnological interest (CIRM-CF; http://www.inra.fr/crb-cirm/), at the National Institute of Agricultural Research (INRA), Marseille, France. The strains were maintained on malt agar slants, using as MA2 (malt extract at 2% w/v) medium for basidiomycetes and MYA2 (malt extract at 2% w/v and yeast extract at 0.1% w/v) medium for ascomycetes and zygomycetes. The T. reesei CL847 secretome (E508 enzymatic cocktail) obtained from IFPEN (Rueil-Malmaison, France) was used as a reference enzymatic cocktail [24, 46].
Culture conditions and secretome preparation
Based on previous studies [22, 23], the fungal cultures were grown in a liquid medium containing 15 g.l-1 (based on the dry matter) of the autoclaved maize bran fraction (provided by ARD, Pomacle, France) as a carbon source, 2.5 g.l-1 of maltose as a starter, 1.842 g.l-1 of diammonium tartrate as a nitrogen source, 0.5 g.l-1 yeast extract, 0.2 g.l-1 KH2PO4, 0.0132 g.l-1 CaCl2.2H2O and 0.5 g.l-1 MgSO4.7H2O. The sugar content (w/w) of the autoclaved maize bran fraction is 16.10% arabinose, 28.73% xylose, 0.17% mannose, 5.65% galactose and 22.06% glucose. Miniaturized fungal cultures were carried out in 16-well baffled plates as described in . The cultures were inoculated with 2 × 105 spores.ml-1 for sporulating fungi, or with mycelial fragments generated using a Fastprep®-24 (MP Biomedicals) set to 5 m.s-1 for 40 s, for non-sporulating fungi prior to incubation in 16-well baffled plates at 30°C with orbital shaking at 140 rpm (Infors HT, Switzerland).
All of the cultures were stopped at 7 days after inoculation. The cultures were stopped at 7 days after inoculation since (i) it is the mid-term growth for most of fungi (basidiomycetes and ascomycetes) in our conditions (16-well baffled plates) and (ii) extended growth led to evaporation of the culture medium. The culture broth (secretome) was harvested (total volume of 20 to 30 ml), filtered (using 0.2 μm polyethersulfone membrane, Vivaspin, Sartorius), diafiltered and concentrated (Vivaspin with a 10 kDa cut-off polyethersulfone membrane, Sartorius) in 50 mM acetate solution buffer, pH 5 to a final volume of 3 ml and then stored in appropriate vials (1.2-ml tubes with septa in cluster plate, ABgene, Thermo scientific, USA) at -20°C until use.
The total amount of protein was assessed using Bradford assays (Bio-Rad Protein Assay Dye Reagent Concentrate, Ivry, France) with a BSA standard that ranged from 0.2 to 1 mg.ml-1.
Saccharification assays on micronised lignocellulosic substrates
The concentrated secretomes were tested for their ability to hydrolyse micronised WS (Triticum aestivum, Apache, France), which was prepared by G. Ghizzi and X. Rouau as described in . WS particles had an average size of 100 μm. A 1% (w/v) WS suspension was prepared in 50 mM acetate buffer, pH 5, supplemented with 40 μg.ml-1 of tetracycline as an antibiotic and 30 μg.ml-1 of cycloheximide as an antifungal agent. The resulting suspension was dispensed into 96-well plates by the Tecan Genesis Evo 200 robot (Tecan, Lyon, France) and the plates were frozen at -20°C until needed.
The saccharification assay was performed using a high-throughput automated method that has been previously described  using a Tecan Genesis Evo 200 robot (Tecan). All of the appropriate blanks and controls reactions were performed as described in . The initial step of wheat-straw hydrolysis was measured after 4 hours using normalized amount of protein (20 μg of T. reesei CL847 enzymatic cocktail and 10 μg of each secretome). The control reaction was performed with 30 μg of T. reesei CL847 enzymatic cocktail. Saccharification was performed at 37°C with 8 Hz shaking. After 4 hours of incubation, the saccharification reactions were filtered and recovered; the reducing sugars were quantified using the DNS method and the glucose content was measured using a Glucose RTU kit (Biomérieux, Marcy l'Etoile, France) following the manufacturer's instructions. All of the reactions were performed independently at least six times.
To quantify the sugars released at the saccharification plateau (24 h for WS), 15 μl of concentrated secretome (5 to 30 μg total proteins) was added to the substrate plate either alone or in addition to 30 μg of the T. reesei CL847 enzymatic cocktail. The released reducing sugars were measured by DNS assay. The saccharification reactions were also analysed by high performance anion exchange chromatography (Dionex, Voisins-le-Bretonneux, France) and a CarboPac PA-1 column (Dionex) to quantify the amounts of glucose and xylose, as described in . All of the reactions were independently performed at least in triplicate.
Enzyme activity measurements
pNP-based chromogenic substrates and complex substrates were used to assay the enzymatic activities of the fungal secretomes. pGlc, pLac, pCel, pXyl, pAra, pGal, pMan and pAc were obtained from Sigma (Sigma-Aldrich, St.Louis, MO). For all pNPs except for pAc, a 1 mM pNP solutions was prepared in 50 mM sodium acetate buffer pH 5, and distributed into polystyrene 96-well plates (IWAKI, Japan) at 100 μl per well and one column per substrate. pNP standards ranging from 0 to 0.2 mM were then added to the plates. The plates were frozen at -20°C until use. The assay was performed by adding appropriate dilutions of 20 μl of the secretome to pNP plates that had been previously preincubated at 37°C. The plates were sealed using a PlateLoc plate sealer (Velocity 11, Agilent) to prevent evaporation, and incubated at 37°C with shaking at 1000 rpm (Mixmate, Eppendorf). After 30 minutes, the reaction was stopped by the addition of 130 μl of 1 M Na2CO3 solution pH 11.5. The amount of pNP released was measured at 410 nm and quantified using a pNP standard curve ranging from 0 to 20 nmol. As pAc is unstable at room temperature in acetate buffer, pH 5, the stock solution of pAc was prepared at 20 mM in DMSO and diluted to 1 mM in 50 mM sodium phosphate, pH 6.5, just prior to use. We immediately added 15 μl of the secretome, and the kinetics of hydrolysis were followed at 410 nm for one minute. One unit of enzyme was defined as 1 μmol of p-nitrophenyl released per mg of enzyme per minute under our experimental conditions.
The complex substrates used in this study were carboxymethyl cellulose (CMC, Sigma), Avicel PH101 (Fluka), birchwood xylan (Sigma), low viscosity wheat arabinoxylan (Megazyme, Wicklow, Ireland), insoluble wheat arabinoxylan (Megazyme), insoluble ivory nut mannan (MAN, Megazyme), carob galactomannan (GMA, Megazyme), larch wood arabinogalactan (Megazyme) and sugar beet arabinan (Megazyme). A 1% w/v suspension/solution was prepared in 50 mM sodium acetate buffer, pH 5, and distributed into 96-well plates at 100 μl per well and one column per substrate. Glucose standards that ranged from 0 to 20 mM were added to the plate. The plates were then frozen at -20°C until use. The assay was performed by adding 20 μl of the secretome (at three appropriate dilutions) to the previously thawed substrate plates. The plates were shaken and incubated at 37°C for one hour using the Tecan Genesis Evo 200 robot plate incubator (Tecan France, Lyon, France), and the reducing sugars were quantified using the automated DNS method, as described in . One unit of enzyme was defined as 1 μmol of glucose equivalent released per mg of enzyme per minute under the assay conditions used here.
The global cellulase activity was determined on filter paper (Whatmann n°1) using 6 mm diameter discs using a protocol adapted from . Vials containing one filter paper disc in 100 μl of 50 mM sodium acetate buffer, pH5, and 50 μl of the secretome were incubated for 2 hours at 50°C. All of the assays were performed in triplicate. After incubation, the reducing sugars were quantified using the automated DNS method  with glucose standards that ranged from 0.2 to 20 mM. One unit of enzyme was defined as 1 μmol of glucose equivalent released per mg of enzyme per minute under our experimental conditions.
Data from the activities measured in all the secretomes except T. reesei CL847 were ordered and grouped using the same clustering method as described for the genome analysis of CAZymes.
Identification of proteins by LC-MS/MS analysis
Proteins from the diafiltered supernatants of A. niger, A. nidulans, U. maydis and T. reesei CL847 (25 μg) were separated by one-dimensional (1D) electrophoresis (Precast Tris-Glycine 12% SDS-PAGE gels, BioRad) and stained with Coomassie blue (BioRad). Each 1D electrophoresis lane was cut into 24 gel fragments (2 mm in width) and protein identification was performed using the PAPPSO platform facilities (http://pappso.inra.fr). In-gel digestion was carried out with the Progest system (Genomic Solution) according to a standard trypsinolysis protocol. Gel pieces were first washed twice with 10% (v/v) acetic acid followed by a wash with 40% (v/v) ethanol in water and then with acetonitrile. They were then further washed with 25 mM NH4CO3 and dehydrated in acetonitrile (two alternating cycles). Digestion was performed for 6 hours at 37°C with 125 ng of modified trypsin (Promega) dissolved in 20% (v/v) methanol in 20 mM NH4CO3. The tryptic peptides were first extracted with 50% (v/v) acetonitrile and 0.5% trifluoroacetic acid in water and then with pure acetonitrile. The two peptide extracts were pooled, dried in a vacuum speed concentrator and suspended in 25 μl of 2% (v/v) acetonitrile and 0.08% (v/v) trifluoroacetic acid in water.
LC-MS/MS analysis was performed on an Ultimate 3000 LC system (Dionex) connected to an LTQ-Orbitrap Discovery mass spectrometer (Thermo Fisher, USA) using a nanoelectrospray ion source. After 4 min, the precolumn was connected to the separating nanocolumn Pepmap C18 (0.075 × 15 cm, 100Å, 3 μm) and the linear gradient was started from 2 to 36% of buffer B (80% acetonitrile and 0.1% formic acid) in buffer A (2% acetonitrile and 0.1% formic acid) at 300 nl.min-1 for 50 min. The doubly and triply charged precursor ions were subjected to MS/MS fragmentation with a 1-min exclusion window and classical peptide fragmentation parameters (Qz = 0.22, activation time = 50 ms, collision energy = 35%).
The raw mass data were first converted to the mzXML format with the ReAdW software (http://tools.proteomecenter.org/software.php). Protein identification was performed by querying the MS/MS data against the corresponding protein databases (UniProtKB, 2009.06.08) along with an in-house contaminant database, using the X!Tandem software (X!Tandem Tornado 2008.02.01.3, http://www.thegpm.org) with the following parameters: one missed trypsin cleavage, alkylation of cysteine and conditional oxidation of methionine precursor and fragment ion set to 10 ppm and 0.5 Da, respectively. A refined search was added with similar parameters, except that the semi-tryptic peptides and the possibly N-terminal acetylated proteins were included. All the peptides that matched with an E-value lower than 0.05 were parsed with an in-house programme (http://http:/PAPPSO.inra.fr/bioinformatique.html). Proteins identified by at least two unique peptides and a log (E-value) lower than 1.E-8 were considered to be validated. A list of proteins and peptides with masses and log (E-value) is provided for T. reesei, A. niger, A. nidulans and U. maydis proteomes in additional files 3, 4, 5 and 6 (Tables S3, S4, S5, S6), respectively.