Global Phosphoproteomic Analysis Reveals the Involvement of Phosphorylation in Aflatoxins Biosynthesis in the Pathogenic Fungus Aspergillus flavus

Aspergillus flavus is a pathogenic fungus that produces toxic and carcinogenic aflatoxins and is the causative agent of aflatoxicosis. A growing body of evidence indicates that reversible phosphorylation plays important roles in regulating diverse functions in this pathogen. However, only a few phosphoproteins of this fungus have been identified, which hampers our understanding of the roles of phosphorylation in A. flavus. So we performed a global and site-specific phosphoproteomic analysis of A. flavus. A total of 598 high-confidence phosphorylation sites were identified in 283 phosphoproteins. The identified phosphoproteins were involved in various biological processes, including signal transduction and aflatoxins biosynthesis. Five identified phosphoproteins associated with MAPK signal transduction and aflatoxins biosynthesis were validated by immunoblotting using phospho-specific antibodies. Further functional studies revealed that phosphorylation of the MAP kinase kinase kinase Ste11 affected aflatoxins biosynthesis in A. flavus. Our data represent the results of the first global survey of protein phosphorylation in A. flavus and reveal previously unappreciated roles for phosphorylation in the regulation of aflatoxins production. The generated dataset can serve as an important resource for the functional analysis of protein phosphorylation in A. flavus and facilitate the elucidation of phosphorylated signaling networks in this pathogen.

equilibrate the column for the last 9.9 min. 23

Raw Data Analysis 24
The raw files from the AB SCIEX Triple TOF™ 5600 Mass Spectrometer were processed 25 with ProteinPilot™ Software v.4.5 Beta and Scaffold, coupled with Mascot. The raw data 26 files (.wiff) were converted into peak lists (mascot generic format; .mgf) and searched, using 27 ProteinPilot™ Software, against A. flavus NRRL 3357 protein database (downloaded from 28 http://www.ncbi.nlm.nih.gov/protein?LinkName=genome_protein&from_uid=360 in January 29 2013 and contained 13,485 protein sequences) concatenated with a reverse decoy database 30 and protein sequences of common contaminants. Enzyme specifity was set as full cleavage by 31 trypsin. Carbamidomethylation (Cys) was set as a fixed modification, whereas alkylation and 32 phosphorylation were emphasized as special factors; biological modifications were entered as 33 ID focus; and false discovery rate (FDR) analysis was set as <1%. The raw files from the Bruker Ion-Trap Mass Spectrometer were converted to .mgf files 39 with DataAnalysis 4.0 (Bruker Compass; Fremont, CA, USA). Mascot 2.3 and pFind Studio 40 2.6 were used for further analyses 1-3 . The search criterion was trypsin digestion, with up to 41 two missed cleavages. Fixed modification was carbamidomethylation (Cys), and variable 42 modifications were oxidation (Met) and phosphorylation (S, T, Y). Precursor ion mass 43 tolerance was set as ± 0.4 Da, and fragment ion mass tolerance was set as ± 0.6 Da. The 44 threshold for the Mascot score was filtered at 1% FDR. To calculate the FDR, the 45 target-decoy strategy was used, as determined by the equation FDR = 2[n Decoy /(n Decoy + 46 n Target )] 4-6 , and the threshold was determined at 1% FDR. If the E score of pFind was <10 -5 or 47 if the Mascot score was >20, peptides were accepted. All raw data and the converted .mgf 48 files have been uploaded to the publicly accessible database PeptideAtlas (dataset ID 49 PASS00400) (http://www.peptideatlas.org/PASS/PASS00400) 7 . 50

Localization of Phosphorylation Sites 51
All MS/MS spectra of phosphopeptides were manually verified using a previously described 52 method 8,9 . The probabilities of phosphorylation at each potential site were evaluated using the 53 Ascore algorithm in ScaffoldPTM for the results of ProteinPilot 10 . Phosphopeptides with an 54 Ascore ≥19 were accepted as a class one site. An Ascore ≥19 was considered to indicate 55 that the phosphorylation assignment had a 99% chance of being correct. For the remaining 56 phosphopeptides without an Ascore, the phosphopeptides with a minimum of 95% confidence 57 as determined by ProteinPilot were accepted, and the spectra were manually evaluated. In addition, the accuracy of the localization of phosphorylation sites determined by Mascot and 59 pFind was confirmed using probability-based PTM scores [11][12][13] . Phosphorylation sites with a 60 probability of >0.75 were considered class one phosphorylation sites. 61

RFSHTADNAHGYKPPSRPGpSPLR, TQTVQpTPPKTEGPATLDDMKEHIDK and 68
RpSLSLPEVR, respectively. These unmodified synthetic peptides were used as control. 69 Following purification of these antigens, immunization and sampling of the antisera from 70 rabbit were performed by AB clonal, according to standard operating procedures. The 71 specificity of the generated antibodies was determined by the manufacturer using ELISA and 72 western blotting. 73

RNA Extraction 74
Mycelia of WT and all genetically engineered mutants were harvested and grinded with liquid 75 nitrogen, and TRIzol reagent (Biomarker Technologies, Beijing, China) was used to isolate 76 RNA. RNA was purified with the DNase (Thermo) and the first strand cDNA was 77 synthesized with TransScript® All-in-One First-Strand cDNA Synthesis SuperMix (TransGen 78 Biotech, Beijing, China).