RESEARCH ARTICLE

Proteomic analysis of the soil filamentous fungus Aspergillusnidulans exposed to a Roundup formulation at a dose causing nomacroscopic effect: a functional study

Florence Poirier1 & Céline Boursier2 & Robin Mesnage3,4 & Nathalie Oestreicher5,6 &

Valérie Nicolas2 & Christian Vélot4,5,6

Received: 24 May 2017 /Accepted: 13 September 2017# Springer-Verlag GmbH Germany 2017

Abstract Roundup® is a glyphosate-based herbicide (GBH)used worldwide both in agriculture and private gardens. Thus,it constitutes a substantial source of environmental contami-nations, especially for water and soil, and may impact a num-ber of non-target organisms essential for ecosystem balance.The soil filamentous fungus Aspergillus nidulans has beenshown to be highly affected by a commercial formulation ofRoundup® (R450), containing 450 g/L of glyphosate (GLY),at doses far below recommended agricultural application rate.In the present study, we used two-dimensional gel electropho-resis combined to mass spectrometry to analyze proteomicpattern changes in A. nidulans exposed to R450 at a dosecorresponding to the no-observed-adverse-effect level(NOAEL) for macroscopic parameters (31.5 mg/L GLYamong adjuvants). Comparative analysis revealed a total of

82 differentially expressed proteins between control andR450-treated samples, and 85% of them (70) were unambig-uously identified. Their molecular functions were mainlyassigned to cell detoxification and stress response (16%), pro-tein synthesis (14%), amino acid metabolism (13%), glycoly-sis/gluconeogenesis/glycerol metabolism/pentose phosphatepathway (13%) and Krebs TCA cycle/acetyl-CoA synthesis/ATP metabolism (10%). These results bring new insights intothe understanding of the toxicity induced by higher doses ofthis herbicide in the soil model organism A. nidulans. To ourknowledge, this study represents the first evidence of proteinexpression modulation and, thus, possible metabolic distur-bance, in response to an herbicide treatment at a dose that doesnot cause any visible effect. These data are likely to challengethe concept of Bsubstantial equivalence^ when applied toherbicide-tolerant plants.

Keywords Roundup® . NOAEL . Aspergillus nidulans .

Proteomics . Metabolism . Herbicide tolerance . Substantialequivalence

Introduction

The effects of intensive agricultural practices on biogeochem-ical flows have been defined as critical for earth-system func-tioning (Steffen et al. 2015). A common feature of biochem-ical cycles is that soil microorganisms are key agents in themaintenance of soil quality and resilience. Sustainability atthis level can be affected by pesticide application, such as byglyphosate-based herbicides (GBH), which are currently themost widely used pesticides worldwide. GBH are used expo-nentially, especially since 80% of genetically modified (GM)plants commercially grown are designed at least to tolerateRoundup® (James 2011): glyphosate (GLY) represented

Responsible editor: Philippe Garrigues

* Christian Vélotchristian.velot@u-psud.fr

1 Université Paris 13, UFR SMBH, Plateforme PPUP13, 1 rue deChablis, 93017 Bobigny cedex, France

2 UMS-IPSIT, US31 Inserm-UMS3679 CNRS, PlateformesTrans-Prot et d’Imagerie Cellulaire, Université Paris-Sud, Faculté dePharmacie, Tour E1, 5 Rue Jean-Baptiste Clément,92296 Châtenay-Malabry, France

3 Gene Expression and Therapy Group, King’s College London,Faculty of Life Sciences & Medicine, Department of Medical andMolecular Genetics, 8th Floor, Tower Wing, Guy’s Hospital, GreatMaze Pond, SE1 9RT London, UK

4 CRIIGEN, 81 rue Monceau, 75008 Paris, France5 Laboratoire VEAC, Université Paris-Sud, Faculté des Sciences, Bât.

360, Rue du Doyen André Guinier, 91405 Orsay, France6 Pôle Risques MRSH-CNRS, Université de Caen, Esplanade de la

Paix, 14032 Caen, France

Environ Sci Pollut ResDOI 10.1007/s11356-017-0217-6

3.7% of the mass of total herbicide active ingredient applied in1996 in the United States, but 53.5% in 2009 (Coupe andCapel 2015). Such intensive use of GBH aggravates soil ero-sion, undermining soil quality by increasing leaching ofbanned remnant pesticides (Sabatier et al. 2014) and of nitrateand phosphate into the environment (Gaupp-Berghausen et al.2015). In plants, GLY disrupts the shikimate pathway throughinhibition of the 5-enolpyruvylshikimate-3-phosphate syn-thase (EPSPS) enzyme, blocking the synthesis of essentialaromatic amino acids and precursors of other critical aromaticcompounds including plant growth regulators and phyto-alexins (Duke et al. 2003). Besides the fact that shikimatepathway is also found in some microorganisms (Bentley1990), some of the GBH toxic effects may also be due to otheringredients mixed with GLY in the commercial formulationssince these latter are much more toxic than GLYalone, wheth-er it is for human cells or tissues (Mesnage et al. 2015) ormicroorganisms (Braconi et al. 2006; Clair et al. 2012;Lipok et al. 2010; Nicolas et al. 2016; Qiu et al. 2013).Thus, GBH use is also suggested to affect soil quality throughtoxic effects on soil microorganisms such as symbiotic my-corrhizal fungi (Zaller et al. 2014) as well as some Aspergillusspecies (Carranza et al. 2014; Nicolas et al. 2016).

In a previous study (Nicolas et al. 2016), we evaluated thetoxicity of a GBH commercial formulation, Roundup BGTplus^ containing 450 g/L of GLY (R450), on the soil filamen-tous fungus Aspergillus nidulans. This ascomycete fungus isan experimental model organism used for decades in basic andindustrial microbiology research (Martinelli and Kinghorn1994). Because it has been extensively studied, it is a well-characterized organism that provides a relevant marker foragricultural soil health. We found that R450 causes multipleeffects affecting various biological processes at doses far be-low agricultural dilution.

The objective of the present work is to further investigatethese toxicological effects using proteomics, in order to getnew insights into the toxic mechanisms caused by GBH ap-plication. Proteomic differential display is a powerful tool foridentifying proteins and studying global cellular responses toa specific environment (de Arruda Grossklaus et al. 2013;Gillardin et al. 2009; Poirier et al. 2001; Sacheti et al. 2014;Shimizu et al. 2009). Not only this approach based on two-dimensional gel electrophoresis (2-DE) and mass spectrome-try (MS) can reveal alterations for genes that are not regulatedat the transcriptional level, but it also allows to detect a num-ber of protein isoformswith specific activities or functions dueto posttranslational modifications (PTMs). Thus, it is probablythe most informative approach to evidence the direct and in-direct effects of particular environmental variations on ecolog-ically relevant species (Lemos et al. 2010), and especially on asoil fungus such as A. nidulans (Kniemeyer 2011). In severalrecent studies, 2-DE has been used to elucidate the effects ofseveral pesticides such as atrazine, butachlor, 2,4-

dichlorophenoxyacetic acid, pentachlorophenol or glyphosatealone (Ahsan et al. 2008; Fang et al. 2010; Kumari et al. 2009;Teixeira et al. 2005; Thornton et al. 2010). To date, only twoproteomic analyses were conducted to understand the effectsof a commercial formulation of Roundup®: one concerned amaize (Mesnage et al. 2016), and the other was about the liverof rats chronically exposed to an ultra-low dose of Roundup®herbicide (Mesnage et al. 2017). Not only is the present studyrelated to an entirely different type of eukaryotic cells, but italso focuses on the effects of a direct exposure under condi-tions of apparent herbicide tolerance (i.e. in the absence ofgeneral toxic effects at the macroscopic level). It is the reasonwhy we chose to perform this proteomic analysis at a dosecorresponding to a no-observed-adverse-effect level(NOAEL) associated to these macroscopic parameters. Thischoice was all the more relevant since we previously showedthat some of the cellular and metabolic effects caused by R450were still evident at this concentration (Nicolas et al. 2016).

Materials and methods

Chemicals

The herbicide Roundup used in this work (R450) was thecommercial formulation called BGT plus^ (450 g/L of GLY,corresponding to 100%), available on the market (homologa-tion 2020448, Monsanto, Anvers, Belgium).

Protease inhibitor cocktail tablets were from Roche(Mannheim, Germany). Linear 18 cm IPG strips pH 4–7,IPG buffer pH 4–7, 2DQuant kit, urea and thiourea were fromGE Healthcare (Uppsala, Sweden). The acrylamide/bisacrylamide solution was from Euromedex (Mundolsheim,France). The SYPRO Ruby solution was from BioRad(Hercules, CA, USA). CHAPS and EZblue were fromSigma-Aldrich (St Louis, USA). Trypsin Porcine was fromPromega (Madison, WI, USA). Acetonitrile was from Merck(Darmstadt, Germany).

A. nidulans strain, growth conditions and proteinextraction

A. nidulans strain used in this study was CV125 (pabaA1)(Nicolas et al. 2016). Media composition, supplements andbasic growth conditions were as described by Cove (1966).Solid media contained 1.2 or 3% agar. Plates were incubatedat 37 °C, and liquid cultures were carried out at 30 °C in anorbital shaker at 150 rpm. Mycelia for protein extraction weregrown at 30 °C for 15 h in 400 mL minimal medium withfructose (0.1%) as the carbon source and urea (5 mM) as thenitrogen source, in the absence (BControl^) or presence(BR450^) of 0.007% R450 (i.e. containing 31.5 mg/L GLYamong adjuvants). For each treatment (absence and presence

Environ Sci Pollut Res

of R450), duplicate cultures were carried out (two culturesseparately inoculated and grown in parallel). Proteins wereextracted as previously described (Shimizu et al. 2009) withminor modifications. Briefly, mycelia were harvested by fil-tration, washed with chilled sterile water, wrung dry andground under liquid nitrogen into a fine powder using a mortarand a pestle. The ground powder was suspended in four vol-umes of cold acetone (− 20 °C) containing 10% (w/v) tricar-boxylic acid (TCA) and stored overnight at − 20 °C. Aftercentrifugation at 15,000×g for 15 min, the precipitate waswashed twice with cold acetone containing 1% (v/v) ß-mercaptoethanol and air-dried for 5 min at room temperature.The pellet was then dissolved in a solution containing 7 Murea, 2M thiourea, 4% (w/v) CHAPS, 0.5% (v/v) Nonidet NP-40 and a protease inhibitor (Roche, Mannheim, Germany) (1tablet for 10 mL). After incubation for 2 h, the lysate wascentrifuged at 150,000×g for 25 min at 4 °C, and the super-natant (protein sample) was stored at − 20 °C. Protein concen-trations were determined using the 2D Quant kit (GEHealthcare).

Two-dimensional gel electrophoresis

The protein samples (100 μg) were added to a solution con-taining 7 M urea, 2 M thiourea, 3% (w/v) CHAPS, 1% (w/v)NP-40, 0.5% (v/v) IPG buffer pH 4–7 and bromophenol blue,to a final volume of 350 μL. After swelling of dry IPG stripswith samples (9 h at 19 °C), focusing was carried out for a totalof 47,260 Vh (Protean IEF Cell focusing system, BioRad).Prior to second-dimension separation, strips were incubatedfor 15 min in 50 mM Tris-HCl, pH 8.6, containing 6 M urea,2% (w/v) SDS, 1% (w/v) DTT, 30% (v/v) glycerol andbromophenol blue, then re-incubated for 15 min in the samebuffer containing 2.5% (w/v) iodoacetamide instead of DTT.Using home-made gels (12%, 1 mm × 200 mm × 250 mm),electrophoresis was carried out for 30 min at 80 V, 30 min at160 V, then for 3 h and 40 min at 600 V (Ettan DALT6Separation unit, GE Healthcare). After protein fixation (50%ethanol, 8% acetic acid) for 1 h, gels were stained usingSYPRO Ruby for comparative image analysis and Coomassieblue for mass spectrometry analysis. SYPRO Ruby-stainedgels were digitized with the Typhoon™ 9400 (GEHealthcare). Coomassie blue-stained gels were digitized withthe ImageScannerII (GE Healthcare).

Comparative image analysis

SYPRO Ruby images were analyzed with the ImageMaster®2D Platinum v5.0 software (GE Healthcare). Quantitative dif-ferential analysis was performed on a set of four gels persample type (BControl^ or BR450^). After automatic spot de-tection, manual checking and correction, approximately 1170spots were detected per gel. Groups of spots were

automatically created by matching a set of gels with a refer-ence gel. Variations in staining intensity between gels weresubjected to scatter plot analysis. Scatter plots indicate therelationship between the spot values (by searching for thelinear dependence between them) from two gels. Thegoodness-of-fit is given by a correlation coefficient value near1. Only gels showing a correlation coefficient value > 0.85were kept for the comparative 2-DE image analysis leading tothree gels per condition. Only groups of spots present in atleast three gels from the same class were considered for dif-ferential analysis.

Mass spectrometry analysis

Spots were excised from the Coomassie blue-stained 2-DEgels, and proteins were in-gel-digested with sequence-gradetrypsin (Promega, Madison, WI, USA) as previously de-scribed (Lescuyer et al. 2003). NanoLC-MS/MS analysiswas performed using an Agilent 1200 Series capillary LCsystem (Agilent Technologies, Palo Alto, USA) coupled to a6330 Ion Trap equipped with the nanospray Chip Cube ionsource (Agilent Technologies). Chromatographic separationof 7 μL injection volume per sample was carried out on aC18 reverse-phase column (Zorbax 300SB-C18, 75 μm,Agilent Technologies). Gradient elution (Froidevaux-Klipfelet al. 2011) was applied for solvents at a flow rate of 200 nL/min. MS/MS spectra were analyzed using the Data Analysissoftware v6.1 (Agilent Technologies). The peptide mass pro-files were analyzed using the MASCOT software (MatrixScience, London, UK). The search parameters used were asfollows: NCBI database; fungi taxonomy; fixed carbamido-methylation for cysteine residues; variable oxidation for me-thionine residues; variable phosphorylation for serine, tyro-sine and threonine residues; peptide mass tolerance of100 ppm (MS) and 0.5 Da (MS/MS) maximum was allowed;only one enzymatic missed cleavage; trypsin as digestioncompound. Validated proteins had at least three differentmatched peptides and a minimum Mascot score of 60.Results leading to Bhypothetical proteins^ were then submit-ted to the Broad Institute database (https://www.broadinstitute.org) as well as to the Aspergillus genomedatabase (http://www.aspergillusgenome.org; Cerqueira et al.2014) to find the name of the gene product.

Microscopy imaging

The A. nidulans strain CV125 was grown on an agar-richmedium surface in the absence or presence of the indicatedR450 dose, and plates were inoculated at 37 °C for 3 days.Imaging was performed according to the inverted agar methodas previously described (Hickey et al. 2005). Differential in-terferential contrast (DIC) images of hyphae apices were ob-tained with an inverted confocal laser scanning microscope

Environ Sci Pollut Res

LSM 510-Meta (Carl Zeiss, Germany), using fluorescentchannel (543 nm laser wavelength) and a transmissiondetector.

Statistical analysis

For each treatment (BControl^ and BR450^), protein extractsfrom each one of the duplicate cultures were analyzed in trip-licate (3 different 2D gel electrophoresis experiments weremade per culture material). Variations in protein abundancewere calculated as the ratio of average values (intensities)for a given spot between the two classes (BControl^ andBR450^). Differential expression was considered significantfor spots with ratio > 1.5. The minimal ratio 1.5, correspond-ing to experimental variations, was defined after quantitativeimage analysis performed between 2-DE gels from the sameclass. Statistically significant differences were estimated bycomparing the empirical distribution functions of these twoclasses. Data points were sorted in ascending order, and thedifferences were considered significant when the empiricaldistributions of the two classes had no overlap at all. Thiscorresponds to a Kolmogorov–Smirnov test with a D valueof 1.

Results and discussion

As part of our previous study on the toxicity of R450 onA. nidulans, we determined the lowest- and no-observed-adverse-effect levels (LOAEL and NOAEL) associated tomacroscopic parameters (Nicolas et al. 2016). The determina-tion of these specific doses was based on growth characteris-tics (growth rate, germination lag time and ratio) andmorphol-ogy (mycelial organization, pigmentation) both in solid andliquid media (Fig. 1a, b). Results obtained at the median lethaldose (LD50) are also shown as a positive control of the growthand morphology effects of R450. Given that mycelial disor-ganization visible at the macroscopic level (Fig. 1b, LD50)was due to abnormal hyphal branch formation (observed atthe microscopic level), we also carried out DIC images (Fig.1c) to ensure that the hyphal branching was normal or aber-rant. All the proteomic analyses described in the present studywere performed at the NOAEL dose (0.007%, i.e. containing31.5 mg/L GLY among adjuvants) associated to macroscopicparameters.

Protein-pattern changes following R450 exposure

Exposure of A. nidulans to the GBH R450 at a concentrationcausing nomorphology and growth effect changed the proteinexpression profile, as revealed by a proteomics approachbased on 2-DE and mass spectrometry to identify proteins thatare differentially regulated. The average numbers of protein

spots detected in untreated and R450-treated samples were1207 and 1167, respectively (Fig. 2a, b). In order to analyzegel similarities or experimental problems (SYPRO Rubystaining, differences in protein loading or image acquisitionproblems), variations in staining intensity between gels weresubjected to scatter plot analysis (Fig. 2c). Fifty spots weremodulated by R450: 32 increased in abundance, while 18decreased. Two examples of protein spots affected by the her-bicide are presented on Fig. 3. Spots of interest were thensubmitted to LC-MS/MS analysis. Fifty up-regulated proteins(including 14 with a ratio of 100, i.e. detected only in thetreated sample) and 32 repressed proteins (including 7 with aratio of 100, i.e. detected only in the control sample) wereunambiguously identi f ied (Table 1 and Table 2,respectively). The functional classification of the putative pro-teins showed that R450 affected several cell metabolic path-ways, mainly detoxification, protein synthesis, amino acidmetabolism, Krebs TCA cycle and acetyl-CoA synthesis, gly-colysis/neoglucogenesis, pentose phosphate pathway andglycerol metabolism (Tables 1 and 2 and Fig. 4).

Main cellular processes modulated during the Roundupexposure

Intermediary metabolism and related pathways

The main intermediary metabolic pathways affected by R450exposure, and their interconnection, are summarized in Fig. 5.The herbicide Roundup was previously shown to promote astrong decrease in the mitochondrial transmembrane potential(Δψ), resulting into an uncoupling of oxidative phosphoryla-tions (Peixoto 2005).We have recently shown that exposure toR450 resulted into a stimulation of TCA cycle enzyme activ-ities in A. nidulans (Nicolas et al. 2016), likely as a result oftheΔψ collapse, the uncoupling effect causing an accelerationof respiration and Krebs TCA cycle. Our present proteomicdata confirmed this observation, since levels of at least twoTCA cycle enzymes (aconitase and α-ketoglutarate dehydro-genase) were increased under R450 exposure (Table 1). Sucha metabolic stimulation may seem inconsistent with the down-regulation of all the acetyl-CoA-producing pathways.However, pyruvate dehydrogenase and the putativemethylmalonate-semialdehyde dehydrogenase (MSDH) weredown-regulated by a factor 2 only, while ATP citrate lyase,whose activity would result into a depletion of mitochondrialcitrate, and consequently into a block of the Krebs TCA cycle,was suppressed (Table 2). In A. nidulans, loss of ATP citratelyase greatly affects growth on carbon sources that do notdirectly result in cytoplasmic acetyl-CoA, such as sugars(Hynes and Murray 2010). This severe phenotype indicatesan absence of compensation by the MSDH pathway, possiblydue to the fact that this branched-chain amino acid (BCAA)degradation pathway would be localized in mitochondria and/

Environ Sci Pollut Res

or peroxisomes as it seems to be the case for plants (Binder2010). Now, the fact that the R450 concentration used in thisstudy did not lead to any growth effect (NOAEL dose associ-ated to growth and morphology), including with fructose as

the carbon source (Nicolas et al. 2016), suggests either thepresence of residual ATP citrate lyase activity (i.e. low levelsof the enzyme under the detection limit on the 2D gels) or aninduction of acetyl-CoA synthetase activity, although this

a

b

c

Fig. 1 Behavior of the fungus A. nidulans when exposed to R450 at theNOAEL, LOAEL and LD50 doses. These doses were 0.007, 0.0075 and0.025% (i.e. containing 31.5, 33.75 and 112.5 mg/L GLY amongadjuvants, respectively). a Plates containing rich medium with theindicated R450 concentration were inoculated with spore of the CV125strain at the central point. b Enlarged view of a sector of the colony shownin (a) to visualize the mycelial organization. c Effects of R450 on hyphae

branching (cell polarity). Images were obtained by differentialinterferential contrast (DIC) microscopy of hyphae apices of the CV125A. nidulans strain grown for 3 days at 37 °C on solid rich media in theabsence or presence of the indicated R450 dose. For each dose, the imagewas representative of observations made on 10 hyphae from each one ofthree independent experiments (n = 30)

Fig. 2 2-DE gels of proteins extracted from A. nidulans grown in theabsence (a) or presence (b) of 0.007%R450. Proteins were separated on alinear pH 4–7 IPG strip, followed by a 12% SDS polyacrylamide gel, asstated in the BMaterials and methods^ section. The gels were stained with

SYPRO Ruby. c Scatter plot result between the spot values (intensities)from two gels (control and R450). Regression line equation is given at thebottom right; Corr: correlation coefficient value; Count: number of iden-tical spots between the two gels

Environ Sci Pollut Res

enzymewas not identified as up-regulated protein under R450exposure at this NOAEL dose (Table 1). Indeed, the other wayof producing cytoplasmic acetyl-CoA in A. nidulans is acetyl-CoA synthetase from acetate. However, in the presence offructose as the sole carbon source, the only source of acetateis pyruvate via pyruvate decarboxylase, which was not detect-ed as up-regulated protein.

The up-regulation of both ketol-acid reductoisomerase and3-isopropylmalate dehydratase combined with the down-regulation of the MSDH pathway would result in an accumu-lation of the three BCAA, valine, leucine and isoleucine. Aprevious study (Shimizu et al. 2010) demonstrated that up-regulated BCCA synthesis in A. nidulans cells functions asan electron sink for regeneration of NAD+ and NADP+ underconditions that result in higher ratios of NAD(P)H/NAD(P)+,such as hypoxia. However, the stimulation in A. nidulans ofthe TCA cycle, resulting from R450 exposure (Nicolas et al.2016), together with the Δψ collapse (i.e. the uncoupling ef-fect) promoted by this herbicide (Peixoto 2005), would alsoresult in an intracellular accumulation of NADH andNADPH.Then, the up-regulation of the BCCA synthesis would be acellular response in A. nidulans to offset the effects ofRoundup, thus preventing an unbalanced NAD(P)H/NAD(P)+ ratio that in turn impairs cellular metabolism.

In A. nidulans, gluconeogenesis is subjected to carbon ca-tabolite repression (Hynes et al. 2007), which is ultimatelymediated by the transcriptional repressor CreA (Félenbokand Kelly 1996). This repression exerts optimally in the pres-ence of glucose and to a lower extent in the presence of fruc-tose (Flipphi et al. 2003). In this study, mycelia for proteinextraction were grown (in the absence or presence of R450)with fructose as the sole carbon source. In such conditions, it

is therefore not surprising to detect the presence of thegluconeogenesis-specific enzymes. This was the case for, atleast, two of them since pyruvate carboxylase and fructose1,6-bisphosphatase proved to be up-regulated in protein ex-pression abundance (with a ratio about 2) under R450 expo-sure relative to the control (Fig. 5 and Table 1). This up-regulation was also observed for enolase (Fig. 5 andTable 1), one of the reversible enzymes that are essential forboth glycolysis and gluconeogenesis. A. nidulans enolase isencoded by acuN, whose expression is transcriptionally acti-vated in response to both gluconeogenic and glycolytic carbonsources, but according to two different and independent ways(Hynes et al. 2007). Now, the two other gluconeogenesis-specific enzymes, phosphoenolpyruvate carboxykinase(PEPCK) and putative glucose-6 phosphatase, were not foundto be up-regulated. However, the up-regulation of NADP-glutamate dehydrogenase combined with down-regulation ofN-acetylglutamate kinase (Fig. 5) suggests a possible accumu-lation of glutamate, which has been shown, in A. nidulans, tobe an inducer of the PEPCK activity (Kelly and Hynes 1981).The absence of up-regulation of a putative glucose-6 phospha-tase implies that such a stimulation of gluconeogenesis wouldlead to production of glucose-6 phosphate, likely to fuel thepentose phosphate pathway (PPP) (Fig. 5), in order to supplythe cell with ribose-5-phosphate for nucleotide synthesis andto produce NADPH. This cofactor is especially required forglutathione reductase to form reduced glutathione (GSH) thatis essential to maintain the redox balance of the cell and toeliminate the xenobiotics (see the BDetoxification pathways^section). The global down-regulation of the PPP (Fig. 5)would avoid a too strong stimulation of the gluconeogenesiswhich, when running simultaneously with glycolysis, would

Fig. 3 Example of differential analysis for down-regulated (a) and up-regulated (b) polypeptides following A. nidulans exposure to 0.007%R450. Histograms on the left, imported from ImageMaster 2D Platinum5.0, show the intensity values of spots considered for analysis: spot 2010(a) corresponding to subunit α of translation initiation factor eIF-1(AN3156) and to transaldolase (AN0240) (see Table 2), and spot 2036(b) corresponding to glyceraldehyde 3-phosphate dehydrogenase

(AN8041) and to an unknown function protein (AN9194) (see Table 1).Three gels were analyzed for BControl^ (histogram columns a–c) andBR450-treated^ (histogram columns d–f) samples. The horizontal purpleline indicates in each case the median measure. Partial 2-DE images froma BControl^ (top) and a BR450^ (down) gel are shown for correspondingspots. Arrows indicate the spots modulated by the herbicide exposure

Environ Sci Pollut Res

Table 1 Proteins up-regulated in A. nidulans after exposure to 0.007% R450 (NOAEL dose)

Gene codea Protein (EC number) FCb M. scorec Mat. pepd Spot number

Amino acid biosynthesis

AN5886 (luA)AN6521 (lysF)AN2526AN4376 (gdhA)AN4290

3-Isopropylmalate dehydratase (4.2.1.33)Homoaconitase (4.2.1.36)Ketol-acid reductoisomerase (1.1.1.86)NADP-glutamate dehydrogenase (1.4.1.4)Methylthioribose-1-P isomerase (5.3.1.23)

1001001.862.71100

1801509312274

76553

10161016182330192111

Protein biosynthesis

AN4550 (dps1)AN1913AN6563AN6563AN1084AN6700AN6700AN1954

Aspartyl-tRNA synthetase (6.1.1.12)Lysil-tRNA synthetase (6.1.1.6)Translation elongation factor eEF-1 subunit γTranslation elongation factor eEF-1 subunit γMitochondrial elongation factor TuTranslation elongation factor eEF-3Translation elongation factor eEF-3Mitoch. ribosomal protein (small subunit)

1001.852.041001002.041.781.95

110868199162153896269

6964135197

12509931333135321116926951560

Glycolysis/gluconeogenesis

AN5746 (acuN)AN8041 (gpdA)AN2875 (fbaA)AN5604 (acuG)AN4462 (pycA)AN4462 (pycA)

Enolase (4.2.1.11)Glyceraldehyde 3-P dehydrogenase (1.2.1.12)Fructose bisphosphate aldolase (4.1.2.13)Fructose 1,6-bisphosphatase (3.1.3.11)Pyruvate carboxylase (6.4.1.1)Pyruvate carboxylase (6.4.1.1)

1.951.761001.861.651.79

11277115105178171

33591817

1560203622341823666668

Krebs TCA cycle

AN5571 (kgdA)AN5525 (acoA)

α-Ketoglutarate dehydrogenase (1.2.4.2)Aconitase (4.2.1.3)

2.181.87

8073

43

29253022

Detoxification/stress response

AN6840AN6840AN5831AN8637 (catA)AN5129 (hsp70)AN12473 (hscA)AN9124

Hydroxyacyl glutathione hydrolase (3.1.2.6)Hydroxyacyl glutathione hydrolase (3.1.2.6)Glutathione S-transferase (2.5.1.18)Catalase A (1.11.1.6)Heat shock protein 70Heat shock protein 70Heat shock protein

1001003.451.972.182.181.63

131191217186300459134

35138202513

262834763056884105710572903

Proteolysis

AN5121AN4557

26S proteasome regulatory particle subunitMitochondrial inner membrane AAA protease

1.861.87

13086

710

18233022

Glycerol metabolism

AN6792 (gfdB) Glycerol-3-phosphate dehydrogenase (1.1.1.8) 5.00 179 11 2959

Purine/pyrimidine biosynthesis

AN6157 (pyrG)AN3626 (adD/ad3)

Orotidine 5′-P decarboxylase (4.1.1.23)P-Ribosylaminoimidazole carboxylase (4.1.1.21)

1003.59

17771

113

26283055

ATP metabolism

AN1211AN8674

Vacuolar ATP synthase subunit HV-type H+-transporting ATPase subunit E

2.88100

296320

64

29423476

Nitrogen/amino acid catabolism

AN10079 (ureB)AN7641AN1808

Urease (3.5.1.5)Peroxisomal copper amine oxidase (1.4.3.6)

L-Amino acid oxidase (1.4.3.2)

1.74100100

362111224

61014

75511521250

Protein trafficking

AN0922AN7687AN3594 (sogA)

Coatomer subunit deltaMitoch. outer membrane translocase receptorVacuolar protein sorting-associated protein

2.181.771.63

84261186

6108

10579362903

Cell cycle/signaling/morphogenesis

AN1911AN4163 (cpcB)

GDP-mannose phosphorylase (2.7.7.22)G-Protein beta subunit

1.842.88

233302

125

14932942

Environ Sci Pollut Res

lead to a total dissipation of energy (without possibility that apool of pyruvate produced by glycolysis can be used to fuelthe Krebs TCA cycle). However, the fact that the reversiblenon-oxidative phase of the PPP would be more severelydown-regulated than the reversible oxidative one (ratio of5.83 for abundance of transaldolase versus ratio of 1.71 for6-phosphogluconolactonase; Table 2) would ensure the simul-taneous synthesis of a sufficient pool of ribose and NADPHwhile avoiding the withdrawal of two intermediates fromglycolysis/gluconeogenesis (fructose-6-phosphate and glycer-aldehyde-3-phosphate) to produce only ribose-5-phosphate(Fig. 5) (such a metabolic pattern occurs when the cellularrequirements are greater in ribose than in NADPH).

R450 exposure resulted into the modulation of the abun-dance of the cytosolic glycerol-3-phosphate dehydrogenaseand the NADP-dependent glycerol dehydrogenase, two keyenzymes regulating the metabolism of glycerol, an importantby-product of glycolysis (Gancedo et al. 1986). Whileglycerol-3-phosphate dehydrogenase that catalyzes the firststep of the glycerol synthesis pathway was up-regulated, theNADP-dependent glycerol dehydrogenase involved in thefirst step of the glycerol catabolism was down-regulated, sug-gesting that R450 exposure resulted into an increase in glyc-erol production rate and its accumulation in A. nidulans (Fig.5). In yeast and A. nidulans, such a glycerol accumulation hasbeen shown to be the main response to a hyperosmotic stress,in order to counterbalance the external osmotic pressure(Blomberg and Adler 1989; Nevoigt and Stahl 1997; Redkaret al. 1995). Consistently, we previously demonstrated thatexposure of A. nidulans to R450 (at doses higher than theNOAEL one) resulted in a slight increase of the spore diam-eter as well as of the hyphae width, suggesting a modificationof the wall structure affecting osmoregulation (Nicolas et al.

2016). Moreover, we also observed that R450 exposure atthese doses resulted into a high degree of hyphae branchingwith multiple, random secondary germ tube emergence at thetip, giving a Bstump^ appearance. Such a disruption of hyphalpolarity could be also a consequence of osmoregulation dis-turbance due, for instance, to a side effect of the surfactantethoxylated etheralkylamine, present in some Roundup for-mulations, including R450, and necessary for an effective up-take of GLY in plants (Riechers et al. 1994). Interestingly, novariation in glycerol metabolism enzyme levels was reportedin the yeast Saccharomyces cerevisiae when exposed to an-other GBH such as Silglif (Braconi et al. 2008). This supportsthe hypothesis that the observed osmotic stress in A. nidulansin response to R450 would be due to one or more adjuvant(s)rather than to GLY itself.

Protein synthesis

About 14% of A. nidulans proteins that are modulated underR450 exposure are relative to protein synthesis (Fig. 4). This isthe case for the translation elongation factors EF1 and EF3and some aminoacyl-tRNA synthetases corresponding alto-gether to six distinct spots up-regulated by R450 (Table 1).These results suggest that protein translation was significantlystimulated when A. nidulans was exposed to the herbicide,although some other spots corresponding to ribosomal pro-teins 40S and 60S and elongation factor EF2 were slightlydown-regulated by R450 (Table 2). In plants, the mode ofaction of GBH is quite clear since GLY annihilates proteinsynthesis by disrupting the shikimic acid pathway throughinhibition of the EPSPS enzyme, thus blocking the synthesisof essential aromatic amino acids (Duke et al. 2003).However, in other organisms, that may not have the shikimate

Table 1 (continued)

Gene codea Protein (EC number) FCb M. scorec Mat. pepd Spot number

AN2412 (cmkA)AN9085AN3739 (snxA)

Calmodulin-dependent protein kinaseU5 snRNP complex subunitRNP domain protein

2.881002.71

385114121

698

294222343019

Uncharacterized/poorly characterized

AN2572AN2731AN2939AN2343AN4171AN8764AN4848AN9194 (cetL)

Dipeptidyl-peptidase (3.4.14.-)Heat shock proteinMitochondrial inner membrane proteinNitroreductase family proteinUnknown function proteinUnknown function proteinUnknown function proteinUnknown function protein

2.555.003.592.913.5910051.76

1656719115658101121154

10611934314

7902959305529853055262829592036

a The corresponding current A. nidulans gene name is indicated between bracketsb Fold changecMascot scored Number of matched peptides

Environ Sci Pollut Res

Table 2 Proteins down-regulated in A. nidulans after exposure to 0.007% R450 (NOAEL dose)

Gene codea Protein (EC number) FCb M. scorec Mat. pepd Spot number

Amino acid metabolism

AN8770AN8866AN3591AN3593

Acetylglutamate kinase (2.7.2.8)D-3-Phosphoglycerate dehydrogenase (1.1.1.95)Methylmalonate semialdehyde DHe (1.2.1.27)Methylthioribulose-1-P-dehydratase (4.2.1.109)

1001002.092.01

161879396

14554

1498149813522258

Protein biosynthesis

AN6330AN3156AN2734AN2734AN3172

Translation elongation factor eEF-2Translation initiation factor eIF-1 subunit α60S acidic ribosomal protein P060S acidic ribosomal protein P040S ribosomal protein S0

11.435.833.232.101.71

841152586396

39633

15032010207020822237

Acetyl-CoA biosynthesis

AN2435 (aclA)AN9403 (pdhC)

ATP citrate lyase (2.3.3.8)Pyruvate dehydrogenase E1 β subunit (1.2.4.1)

1002.10

97148

76

14982082

Detoxification/stress response

AN9339 (catB)AN7388 (cpeA)AN10220 (ccp1)AN0858 (hsp104)AN4616 (ssz1)

Catalase B (1.11.1.6)Catalase-peroxidaseMitochondrial cytochrome c peroxidaseHeat shock proteinHeat shock protein 70

3.471001.671001.57

122114116265232

4761616

81896721967341109

Glycerol metabolism

AN5563 (gldB)AN5563 (gldB)

NADP(+)-dependent glycerol DHe (1.1.1.72)NADP(+)-dependent glycerol DHe (1.1.1.72)

2.093.47

10898

33

2049818

Pentose phosphate pathway

AN0240 (pppA)AN0285

Transaldolase (2.2.1.2)6-Phosphogluconolactonase (3.1.1.31)

5.831.71

8967

63

20102237

ATP metabolism

AN2315 Mitochondrial ATP-synthase beta chain 11.43 313 7 1503

Cytoskeleton/morphogenesis/cell cycle

AN7570 (tubB)AN0316 (tubA)AN2126AN5686 (tpmA)AN0410 (bimG)

Tubulin alpha-2 chainTubulin alpha-1 chainF-actin capping protein subunit αTropomyosinSerine/threonine protein phosphatase

1.711.712.101003.23

72547811818278

1112383

13711371208225162070

Carbon catabolism

AN7590 NADP-dependent mannitol DHe (1.1.1.138) 2.01 149 10 2258

Vesicle trafficking

AN3416 (ssoA) SNARE-domain containing protein 2.10 96 5 2082

NAD biosynthesis

AN1745 Nicotinamide mononucleotide adenylyl transferase (2.7.7.1) 3.23 86 3 2070

Splicing

AN4978 Pre-RNA splicing factor 100 120 3 3057

Uncharacterized/poorly characterized

AN3996AN8228AN7587AN9194 (cetL)

Methyltransferases family proteinUBX domain-containing proteinUnknown function proteinUnknown function protein

2.1811.431.672.09

776180121

3459

2363150321962049

a The corresponding current A. nidulans gene name is indicated between bracketsb Fold changecMascot scored Number of matched peptidese Dehydrogenase

Environ Sci Pollut Res

pathway, mechanisms responsible for Roundup toxicity ap-pear multiple and may vary from one organism or cell typeto another. In A. nidulans, as in other fungi and bacteria, theshikimate pathway is also present, but that does not necessar-ily imply that GBH operate as in plants. Indeed, there are twoclasses of EPSPS: GLY-sensitive, as in plants (class I) andGLY-tolerant (class II). It is not really clear whether EPSPSfrom A. nidulans belongs to class I or class II, but our previousdata (Nicolas et al. 2016) indicate that GLY does not inhibit(or only partially) the A. nidulans EPSPS enzyme, and that themultiple cellular effects of R450 in this fungus are likely dueto various targets of GLY and/or additives present in the for-mulation. Indeed, we have shown that these additives are notinert since the formulation R450 proved to be much moreactive than GLY alone (Nicolas et al. 2016). Similarly, theantimicrobial effect of glyphosate formulations on some pro-tozoa (patent US7771736 B2) is not necessarily due to an(exclusive) inhibition of the shikimate pathway in these para-sites. We already demonstrated that the mode of action ofRoundup on energetic metabolism in A. nidulans (Nicolaset al. 2016) was different of that previously observed for hu-man cells or tissues (Mesnage et al. 2015; Peixoto 2005).Obviously, the data reported here indicate that the mode ofaction of Roundup on protein synthesis in A. nidulans is to-tally different, even diametrically opposed to that observed inplants. Such a possible stimulation of protein synthesis is con-sistent with the apparent up-regulation of some amino acidbiosynthesis, such as valine, leucine, isoleucine, lysine andglutamate (Table 1 and Fig. 5).

Detoxification pathways

The three detoxification pathways modulated in response toR450 exposure are described in Fig. 6.Methylglyoxal is main-ly a by-product of glycolysis. Its high cytotoxicity implies thatit does not accumulate into the cell. In A. nidulans,methylglyoxal has been shown to be a substrate of NADP-

glycerol dehydrogenase that can reduce it while simultaneous-ly oxidizing NADPH (Schuurink et al. 1990). The down-regulation of this enzyme under R450 exposure (Fig. 5 andTable 2) suggests a possible accumulation of methylglyoxal inthe presence of the herbicide. Another methylglyoxal detoxi-fication pathway involves glyoxalase enzymes (I and II) andreduced GSH to generate pyruvate (Fig. 6a). This pathwayproved to be strongly stimulated in A. nidulans exposed toR450, since herbicide treatment highly increased abundanceof glyoxalase II: the two spots corresponding to this enzymewere not detected in the control 2-DE gels (ratio of 100 forspots 2628 and 3476; Table 1).

Our data reveal that two other detoxification pathways arestimulated in A. nidulans under R450 exposure: one, involv-ing again GSH and a glutathione S-transferase (GST), contrib-utes to xenobiotic resistance (Fig. 6b), and the other, involvingcatalase and peroxidase activities, is required to transform theharmful oxygen compound H2O2 by reducing it to water (Fig.6c). Interestingly, in A. nidulans, the fungal glutathione sys-tem has been shown to function as an antioxidant process andto interplay with the hydrogen peroxide defense mechanism(Sato et al. 2009). Thus, these two pathways are essential inmaintaining cellular redox homeostasis in response to oxida-tive stress. Many studies have shown that GLY induces oxi-dative stress (Ahsan et al. 2008; deAguiar et al. 2016; deMeloTarouco et al. 2016; Gomes et al. 2017; Gomes and Juneau2016; Martini et al. 2016; Mesnage et al. 2015; Murussi et al.2016; Salvio et al. 2016; Wu et al. 2016). The fact that GLYacts as a protonophore (Olorunsogo 1990) increasing mito-chondrial membrane permeability to protons and Ca2+ canexplain such a cellular effect of GLY. Indeed, Ca2+ is consid-ered to be one of the major stimulators of mitochondrial reac-tive oxygen species (ROS) accumulation because it promotesstructural alterations of the inner mitochondrial membrane(Kowaltowski and Vercesi 1999). ROS are highly reactiveand could impair cellular molecules such as lipids, proteinsor DNA, leading to cell damages. Such damages are reflected

14.0

13.0

13.0

10.016.0

34.0

Protein synthesis

Glycolysis/neoglucogenesis/Glycerol metabolism/Pentosephosphate pathwayAmino acid metabolism

TCA cycle / Acetyl CoAsynthesis/ATP metabolism

Detoxification and stress response

Others

Fig. 4 Assignment of identifiedproteins to cellular processes.Spots down- and up-regulatedwere subjected to tryptic digestionand LC-MS/MS analysis. From50 modulated spots, 85% of theproteins (70/82) wereunambiguously identified andassigned to a cellular process ormetabolic pathway

Environ Sci Pollut Res

by aspartate aminotransferase (ASAT) and alanine amino-transferase (ALAT) increased activities in laboratory animalsexposed to GLY (Mesnage et al. 2015).

GST abundance was increased by 3.45-fold under R450treatment (Table 1). GSTs are ubiquitous enzymes, which playa key role in the cell detoxification of a wide range of com-pounds. GSTs catalyze the conjugation of reduced GSH, viaits thiol group, to electrophilic centers of a wide variety ofxenobiotic substrates (X), in order to make these compoundsmore hydrophilic (GS-X), thus facilitating their breakdownand elimination (Fig. 6b).

In A. nidulans, three catalase and/or peroxidase activi-ties have been identified: catalase A, catalase B and

catalase-peroxidase. Only catalase A proved to be up-regulated in response to R450 exposure (Table 1,Fig. 6c), while the two other activities were down-regulated (Table 2, Fig. 6c), especially catalase-peroxidase whose corresponding spot was not detected inthe R450 2-DE gels (Table 2). However, this enzymewould be required mainly during sexual development(Scherer et al. 2002). Now, the difference observed be-tween the two catalases might be due to the fact thatthe gene catA (encoding catalase A) but not catB (encodingcatalase B) is induced by osmotic stress (Kawasaki et al. 1997;Navarro and Aguirre 1998), consistent with the modulation ofglycerol metabolism as described in the previous texts.

α-Ketoglutarate

isocitrate

Citrate

Acétyl-CoA

Malate

Oxaloacétate

α-Ketoglutaratedehydrogenase

Aconitase

ATP citratelyase

Pyruvate

Pyruvatecarboxylase

TCACycle

Glutamate

NADP-Glutamate dehydrogenase

Pyruvatedehydrogenase

Succinyl-CoA

Glucose 6-P

Phosphoenolpyruvate

Malate

Oxaloacétate

esoculGP-6 esotcurF6-Phospho

gluconolactonase

Transaldolase

Ribulose 5-P

Ribose 5-P+

Xylulose-5-P

PentosePhosphatePathway

N-acetylglutamatekinase

IsoleucineValine

Ornithine

Arginine

IsoleucineLeucine

G-3-P dehydrogenase

Fructose 1.6bisphosphatase

Enolase

Fructose 1.6-P2

Aldolase

DHAPGlycerol

Glycerol 3-P

Glycerol-3-Pdehydrogenase

Glycerolmetabolism

Valine

Succinate

Fumarate

Pyruvate

MITOCHONDRION

Fumarate

Succinate

α-Ketoglutarate

Citrate

Acetyl-CoA

MSDH

MSDH

3-isopropylmalate dehydratase

Ketol-acidreductoisomerase

2-P-Glycérate

3-P-Glycérate

1,3-Bisphosphoglycérate

Glyceraldehyde-3-P

DHANADP-Glyceroldehydrogenase

6-phosphogluconolactone

6-phosphogluconate

Fig. 5 Compilation of the mainintermediary metabolism andrelated pathways affected byR450 exposure, and theirinterconnections. Up-regulated(red arrows) and down-regulated(blue arrows) proteins areoutlined in red and blue, respec-tively. Green frames indicatemetabolic pathways whose celllocation is not well-defined and/orsome enzyme activities have notbeen characterized in A. nidulans(the putative functionality of thecognate modulated proteins isbased solely on sequence homol-ogies). MSDH: methylmalonate-semialdehyde dehydrogenase

Environ Sci Pollut Res

Conclusions

Consistent with our previous toxicological data (Nicolas et al.2016), this proteomic analysis evidenced that R450 affectsoxidative metabolism, and especially stimulates Krebs TCAcycle, thus confirming the unexpected impact of this herbicideon the energetic metabolism in this fungus, since Roundupwas previously shown to damage basic mitochondrial func-tions, including the TCA cycle, in mammalian cells (Mesnageet al. 2015; Peixoto 2005). Based on enzyme activitymeasure-ments, such a TCA cycle stimulation was not only observed athigher doses, but also, to a lesser extent, at the concentrationused here, i.e. the NOAEL dose associated to macroscopicparameters (Nicolas et al. 2016). Moreover, the present studyrevealed an alteration of translation functions and amino acidmetabolism that suggested an enhancement of protein synthe-sis. Such data indicate again a different mode of action ofRoundup, when compared to plants or other organisms witha class I EPSPS enzyme. Indeed, in such organisms, proteinsynthesis is inhibited and finally abolished by blocking thesynthesis of essential aromatic amino acids (Duke et al. 2003).

Thus, mechanisms responsible for Roundup toxicity appearmultiple and can vary from one organism or cell type to another.This broad spectrum of effects of Roundup, not only within aspecific organism but also according to the exposed organism, islikely due to various targets of GLYand/or adjuvants present inthe formulations. Indeed, the additives cannot be considered to beinert since many studies have shown that the GBH formulationsare much more toxic than GLYalone (Braconi et al. 2006; Clairet al. 2012; Cuhra et al. 2013; Lipok et al. 2010; Mesnage et al.2013, Mesnage et al. 2014; Mottier et al. 2013; Nicolas et al.2016; Piola et al. 2013; Qiu et al. 2013).

Our data also evidenced the up-regulation of three detoxi-fication pathways, accounting for both osmotic and oxidativestresses. One is necessary for the elimination ofmethylglyoxal, the accumulation of which may result, at leastin part, from the modulation of the glycerol metabolism,which was also revealed in this analysis. The two others arerequired to maintain the redox balance of the cell, and their

stimulation was consistent with the known oxidative stressinduced by GBH, especially because of its uncouplingeffect.

Altogether, this proteomic analysis revealed multiple mo-lecular and metabolic effects of a commercial formulation ofRoundup (R450) in A. nidulans, at a dose that does not affectthe macroscopic parameters. All these metabolic modulationswould be a cellular response to offset the effects of the herbi-cide. Such metabolic adaptations would therefore be sufficientat this dose (0.007%) to prevent the appearance of specificphenotypes (growth and morphology), while they would beonly partial or even inoperative at higher doses (Nicolas et al.2016). Indeed, it has been shown in particular that theuncoupling effect of Roundup is dose-dependent (Peixoto2005). These data imply that metabolic disturbances due topesticide residues may occur at exposure doses for whichthere is no visible toxic effect, regardless of the exposed or-ganism. This is especially the case of agricultural doses forGM plants designed to tolerate herbicides. The assessment ofthese plants is based on the principle of Bsubstantialequivalence^ (i.e. close nutritional and compositional similar-ity between two crop-derived foods) used to claim that GMcrops are as safe and nutritious as currently consumed con-ventional plant-derived foods (Aumaitre et al. 2002).Surprisingly, the presence of herbicide residues in such plantsis ignored (Cuhra 2015). The data reported here confirm, how-ever, the importance of taking into account the impact of theseherbicide residues in the determination process of substantialequivalence, since metabolic disturbances due to these resi-dues may add toxic properties to the final plant product.

Acknowledgments This work was supported by the non-governmentalorganization BGénérations Futures^ and the Committee for IndependentResearch and Information on Genetic Engineering (CRIIGEN), in theframework of a participatory research project.

Funding information It received funding from the Regional Council Ile-de-France and the University Paris-Sud.

Méthylglyoxal

Glutathione

(R)-S-Lactoylglutathione

H2O

D-lactic acidPyruvate

Glutathione

hydroxyacylglutathione hydrolase

(Glyoxalase II)

Glutathione (GSH)

G-S-XXénobiotic (X)

Glutathionetransférase

Elimination

2 H2O2 O2 + 2 H2O

Catalase A

Catalase B

Catalase-peroxidase

cbaFig. 6 Detoxification pathwaysmodulated in response to R450treatment. Methylglyoxal (a),xenobiotics (b) and hydrogenperoxide (c) detoxifying systems.Variation in abundance of R450-responsive proteins is indicatedby the direction of the gray arrows(up or down)

Environ Sci Pollut Res

References

AhsanN, Lee D-G, Lee K-W,Alam I, Lee S-H, Bahk JD, Lee B-H (2008)Glyphosate-induced oxidative stress in rice leaves revealed by pro-teomic approach. Plant Physiol Biochem 46:1062–1070

Aumaitre A, Aulrich K, Chesson A, Flachowsky G, Piva G (2002) Newfeeds from genetically modified plants: substantial equivalence, nu-tritional equivalence and safety for animals and animal products.Livest Prod Sci 74:223–238

Bentley R (1990) The shikimate pathway—a metabolic tree with manybranches. Crit Rev Biochem Mol Biol 25:307–384

Binder S (2010) Branched-chain amino acid metabolism in Arabidopsisthaliana. Arabidopsis Book. https://doi.org/10.1199/tab.0137

Blomberg A, Adler L (1989) Roles of glycerol and glycerol-3-phosphatedehydrogenase (NAD+) in acquired osmotolerance ofSaccharomyces cerevisiae. J Bacteriol 171:1087–1092

Braconi D, Possenti S, Laschi M, Geminiani M, Lusini P, Bernardini G,Santucci A (2008) Oxidative damage mediated by herbicides onyeast cells. J Agric Food Chem 56:3836–3845

Braconi D, Sotgiu M, Millucci L, Paffetti A, Tasso F, Alisi C, Martini S,Rappuoli R, Lusini P, Sprocati AR, Rossi C, Santucci A (2006)Comparative analysis of the effects of locally used herbicides andtheir active ingredients on a wild-type wine Saccharomycescerevisiae strain. J Agric Food Chem 54:3163–3172

Carranza CS, Barberis CL, Chiacchiera SM, Magnoli CE (2014)Influence of the pesticides glyphosate, chlorpyrifos and atrazine ongrowth parameters of nonochratoxigenic Aspergillus section Nigristrains isolated from agricultural soils. J Environ Sci Health 49:747–755

Cerqueira GC, Arnaud MB, Inglis DO, Skrzypek MS, Binkley G,Simison M, Miyasato SR, Binkley J, Orvis J, Shah P, Wymore F,Sherlock G, Wortman JR (2014) The Aspergillus genome database:multispecies curation and incorporation of RNA-Seq data to im-prove structural gene annotations. Nucleic Acids Res 42:705–710

Clair E, Linn L, Travert C, Amiel C, Séralini G-E, Panoff J-M (2012)Effects of Roundup(®) and glyphosate on three food microorgan-isms: Geotrichum candidum, Lactococcus lactis subsp. cremorisand Lactobacillus delbrueckii subsp. bulgaricus. Curr Microbiol64:486–491

Coupe RH, Capel PD (2015) Trends in pesticide use on soybean, corn andcotton since the introduction of major genetically modified crops inthe United States. Pest Manag Sci 72:1013–1022

Cove DJ (1966) The induction and repression of nitrate reductase in thefungus Aspergillus nidulans. Biochim Biophys Acta 113:51–56

Cuhra M (2015) Review of GMO safety assessment studies: glyphosateresidues in Roundup Ready crops is an ignored issue. Env Sci Eur27:20. https://doi.org/10.1186/s12302-015-0052-7

CuhraM, Traavik T, Bøhn T (2013) Clone- and age-dependent toxicity ofa glyphosate commercial formulation and its active ingredient inDaphnia magna. Ecotoxicology 22:251–262

de Aguiar LM, Figueira FH, Gottschalk MS, da Rosa CE (2016)Glyphosate-based herbicide exposure causes antioxidant defenceresponses in the fruit fly Drosophila melanogaster. CompBiochem Physiol C Toxicol Pharmacol 185-186:94–101

de Arruda Grossklaus D, Bailão AM, Vieira Rezende TC, Borges CL, deOliveira MA, Parente JA, de Almeida Soares CM (2013) Responseto oxidative stress in Paracoccidioides yeast cells as determined byproteomic analysis. Microbes Infect 15:347–364

de Melo Tarouco F, de Godoi FG, Velasques RR, da Silveira GA, GeihsMA, da Rosa CE (2016) Effects of the herbicide Roundup on thepolychaeta Laeonereis acuta: cholinesterases and oxidative stress.Ecotoxicol Environ Saf 135:259–266

Duke SO, Baerson SR, Rimando AM (2003) Herbicides: glyphosate. In:Plimmer JR, Gammon DW, Ragsdale NN (eds) Encyclopedia ofagrochemicals. John Wiley & Sons, New York, pp 708–869

Fang Y, Gao X, Zha NB, Li X, Gao Z, Chao F (2010) Identification ofdifferential hepatic proteins in rare minnow (Gobiocypris rarus)exposed to pentachlorophenol (PCP) by proteomic analysis.Toxicol Lett 199:69–79

Félenbok B, Kelly JM (1996) Regulation of carbon metabolism in my-celial fungi. In: Marzluf G, Beambl R (eds) The Mycota III: bio-chemistry and molecular biology, 1st edn. Springer, BerlinHeidelberg New York, pp 369–380

Flipphi M, van de Vondervoort PJ, Ruijter GJ, Visser J, Arst HN Jr,Felenbok B (2003) Onset of carbon catabolite repression inAspergillus nidulans. Parallel involvement of hexokinase and glu-cokinase in sugar signaling. J Biol Chem 278:11849–11857

Froidevaux-Klipfel L, Poirier F, Boursier C, Crépin R, Poüs C, Baudin B,Baillet A (2011) Modulation of septin and molecular motor recruit-ment in the microtubule environment of the Taxol-resistant humanbreast cancer cell line MDA-MB-231. Proteomics 11:3877–3886

Gancedo C, Llobell A, Ribas J-C, Luchi F (1986) Isolation and charac-terization ofmutants from Schyzosaccharomyces pombe defective inglycerol catabolism. Eur J Biochem 159:171–174

Gaupp-Berghausen M, Hofer M, Rewald B, Zaller JG (2015)Glyphosate-based herbicides reduce the activity and reproductionof earthworms and lead to increased soil nutrient concentrations.Sci Rep 5:12886

Gillardin V, Silvestre F, Dieu M, Delaive E, Raes M, Thome J-P,Kestemont P (2009) Protein expression profiling in the Africanclawed frog Xenopus laevis tadpoles exposed to the polychlorinatedbiphenyl mixture Aroclor 1254. Mol Cell Proteomics 8:596–611

Gomes MP, Bicalho EM, Smedbol É, Cruz FV, Lucotte M, Garcia QS(2017) Glyphosate can decrease germination of glyphosate-resistantsoybeans. J Agric Food Chem 65:2279–2286

GomesMP, Juneau P (2016) Oxidative stress in duckweed (LemnaminorL.) induced by glyphosate: is the mitochondrial electron transportchain a target of this herbicide? Environ Pollut 218:402–409

Hickey PC, Swift SR, Roca MG, Read ND (2005) Live-cell imaging offilamentous fungi using vital fluorescent dyes and confocal micros-copy. In: Savidge T, Pothoulakis C (eds) Methods in microbiology,Microbial Imaging, vol 35. Elsevier, London, pp 63–87

HynesMJ, Murray SL (2010) ATP-citrate lyase is required for productionof cytosolic acetyl coenzyme A and development in Aspergillusnidulans. Eukaryot Cell 9:1039–1048

Hynes MJ, Szewczyk E, Murray SL, Suzuki Y, Davis MA, Sealy-LewisHM (2007) Transcriptional control of gluconeogenesis inAspergillus nidulans. Genetics 176:139–150

James C (2011) Global status of commercialized biotech/GM crops.ISAAA brief no 43. ISAAA, Ithaca, NY

Kawasaki L, Wysong D, Diamond R, Aguirre J (1997) Two divergentcatalase genes are differentially regulated during Aspergillusnidulans development and oxidative stress. J Bacteriol 179:3284–3292

Kelly JM, Hynes MJ (1981) The regulation of phosphoenolpyruvatecarboxykinase and the NADP-linked malic enzyme in Aspergillusnidulans. J Gen Microbiol 123:371–375

Kniemeyer O (2011) Proteomics of eukaryotic microorganisms: the med-ically and biotechnologically important fungal genus Aspergillus.Proteomics 11:3232–3243

Kowaltowski AJ, Vercesi AE (1999) Mitochondrial damage induced byconditions of oxidative stress. Free Radic Biol Med 26:463–471

Kumari N, Narayan OP, Rai LC (2009) Understanding butachlor toxicityin Aulosira fertilissima using physiological, biochemical and prote-omic approaches. Chemosphere 77:1501–1507

Lemos MF, Soares AM, Correia AC, Esteves AC (2010) Proteins inecotoxicology—how, why and why not? Proteomics 10:873–887

Environ Sci Pollut Res

Lescuyer P, Strub JM, Luche S, Diemer H, Martinez P, Van Dorsselaer A,Lunardi J, Rabilloud T (2003) Progress in the definition of a refer-ence human mitochondrial proteome. Proteomics 3:157–167

Lipok J, Studnik H, Gruyaert S (2010) The toxicity of Roundup® 360 SLformulation and its main constituents: glyphosate andisopropylamine towards non-target water photoautotrophs.Ecotoxicol Environ Saf 73:1861–1868

Martinelli SD, Kinghorn JR (1994) Aspergillus: 50 years on—progress inindustrial microbiology. Volume 29. Elsevier, Amsterdam—London—New York—Tokyo

Martini CN, Gabrielli M, Brandani JN, Vila Mdel C (2016) Glyphosateinhibits PPAR gamma induction and differentiation of preadipocytesand is able to induce oxidative stress. J Biochem Mol Toxicol 30:404–413

Mesnage R, Agapito-Tenfen SZ, Vilperte V, Renney G, Ward M, SéraliniGE, Nodari RO, Antoniou MN (2016) An integrated multi-omicsanalysis of the NK603 Roundup-tolerant GM maize reveals metab-olism disturbances caused by the transformation process. Sci Rep19:37855

Mesnage R, Bernay B, Séralini GE (2013) Ethoxylated adjuvants ofglyphosate-based herbicides are active principles of human cell tox-icity. Toxicology 313:122–128

Mesnage R, Defarge N, Spiroux de Vendômois J, Séralini GE (2014)Major pesticides are more toxic to human cells than their declaredactive principles. Biomed Res Int. https://doi.org/10.1155/2014/179691

Mesnage R, Defarge N, Spiroux de Vendomois J, Séralini G-E (2015)Potential toxic effects of glyphosate and its commercial formulationsbelow regulatory limits. Food Chem Toxicol 84:133–153

Mesnage R, Renney G, Séralini GE, Ward M, Antoniou MN (2017)Multiomics reveal non-alcoholic fatty liver disease in rats followingchronic exposure to an ultra-low dose of Roundup herbicide. SciRep 9:39328

Mottier A, Kientz-Bouchart V, Serpentini A, Lebel JM, Jha AN, Costil K(2013) Effects of glyphosate-based herbicides on embryo-larval de-velopment and metamorphosis in the Pacific oyster, Crassostreagigas. Aquat Toxicol 128-129:67–78

Murussi CR, Costa MD, Leitemperger JW, Guerra L, Rodrigues CC,Menezes CC, Severo ES, Flores-Lopes F, Salbego J, Loro VL(2016) Exposure to different glyphosate formulations on the oxida-tive and histological status of Rhamdia quelen. Fish PhysiolBiochem 42:445–455

Navarro RE, Aguirre J (1998) Posttranscriptional control mediates celltype-specific localization of catalase A during Aspergillus nidulansdevelopment. J Bacteriol 180:5733–5738

Nevoigt E, Stahl U (1997) Osmoregulation and glycerol metabolism inthe yeast Saccharomyces cerevisiae. FEMS Microbiol Rev 21:231–241

Nicolas V, Oestreicher N, Vélot C (2016) Multiple effects of a commer-cial Roundup® formulation on the soil filamentous fungusAspergillus nidulans at low doses: evidence of an unexpected impacton energetic metabolism. Environ Sci Pollut Res 23:14393–14404

Olorunsogo OO (1990) Modification of the transport of protons and Ca2+

ions across mitochondrial coupl ing membrane by N-(phosphonomethyl)glycine. Toxicology 61:205–209

Peixoto F (2005) Comparative effects of the Roundup and glyphosate onmitochondrial oxidative phosphorylation. Chemosphere 61:1115–1122

Piola L, Fuchs J, Oneto ML, Basack S, Kesten E, Casabé N (2013)Comparative toxicity of two glyphosate-based formulations toEisenia andrei under laboratory conditions. Chemosphere 91:545–551

Poirier F, Pontet M, Labas V, le Caër JP, Sghiouar-Imam N, Raphaël M,Caron M, Joubert-Caron R (2001) Two-dimensional database of aBurkitt lymphoma cell line (DG 75) proteins: protein pattern chang-es following treatment with 5′-azycytidine. Electrophoresis 22:1867–1877

Qiu H, Gen J, Ren H, Xia X, Wang X, Yu Y (2013) Physiological andbiochemical responses ofMicrocystis aeruginosa to glyphosate andits Roundup® formulation. J Hazard Mater 248-249:172–176

Redkar RJ, Locy RD, Singh (1995) Biosynthetic pathways of glycerolaccumulation under salt stress in Aspergillus nidulans. Exp Mycol19:241–246

Riechers DE, Wax LM, Liebl RA, Bush DR (1994) Surfactant-increasedglyphosate uptake into plasma membrane vesicles isolated fromcommon lambsquarters leaves. Plant Physiol 105:1419–1425

Sabatier P, Poulenard J, Fanget B, Reyss JL, Develle AL, Wilhelm B,Ployon E, Pignol C, Naffrechoux E, Dorioz JM, Montuelle B,Arnaud F (2014) Long-term relationships among pesticide applica-tions, mobility, and soil erosion in a vineyard watershed. Proc NatlAcad Sci U S A 111:15647–15652

Sacheti P, Patil R, Dube A, Bhonsle H, Thombre D, Marathe S, VidhateR, Wagh P, Kulkarni M, Rapole S, Gade W (2014) Proteomics ofarsenic stress in the gram-positive organism Exiguobacterium sp. PSNCIM 5463. Appl Microbiol Biotechnol 98:6761–6773

Salvio C, Menone ML, Rafael S, Iturburu FG, Manetti PL (2016)Survival, reproduction, avoidance behavior and oxidative stress bio-markers in the earthworm Octolasion cyaneum exposed to glypho-sate. Bull Environ Contam Toxicol 96:14–319

Sato I, Shimizu M, Hoshino T, Takaya N (2009) The glutathione systemof Aspergillus nidulans involves a fungus-specific glutathione S-transferase. J Biol Chem 284:8042–8053

Scherer M, Wei H, Liese R, Fischer R (2002) Aspergillus nidulanscatalase-peroxidase gene (cpeA) is transcriptionally induced duringsexual development through the transcription factor StuA. EukaryotCell 1:725–735

Schuurink R, Busink R, Hondmann DH, Witteveen CF, Visser J (1990)Purification and properties of NADP(+)-dependent glycerol dehy-drogenases from Aspergillus nidulans and A. niger. J Gen Microbiol136:1043–1050

Shimizu M, Fujii T, Masuo S, Fujita K, Takaya N (2009) Proteomicanalysis of Aspergillus nidulans cultured under hypoxic conditions.Proteomics 9:7–19

Shimizu M, Fujii T, Masuo S, Takaya N (2010) Mechanism of de novobranched-chain amino acid synthesis as an alternative electron sinkin hypoxic Aspergillus nidulans cells. Appl Environ Microbiol 76:1507–1515

SteffenW, RichardsonK, Rockström J, Cornell SE, Fetzer I, Bennett EM,Biggs R, Carpenter SR, de Vries W, de Wit CA, Folke C, Gerten D,Heinke J, Mace GM, Persson LM, Ramanathan V, Reyers B, SörlinS (2015) Sustainability. Planetary boundaries: guiding human devel-opment on a changing planet. Science 347:1259855

Teixeira MC, Santos PM, Fernandes AR, Sá-Correia I (2005) A proteomeanalys is of the yeast response to the herbic ide 2,4-dichlorophenoxyacetic acid. Proteomics 5:1889–1901

Thornton BJ, Elthon TE, Cerny RL, Siegfried BD (2010) Proteomicanalysis of atrazine exposure in Drosophila melanogaster(Diptera: Drosophilidae). Chemosphere 81:235–241

Wu L, Qiu Z, Zhou Y, Du Y, Liu C, Ye J, Hu X (2016) Physiologicaleffects of the herbicide glyphosate on the cyanobacteriumMicrocystis aeruginosa. Aquat Toxicol 178:72–79

Zaller JG, Heigl F, Ruess L, Grabmaier A (2014) Glyphosate herbicideaffects belowground interactions between earthworms and symbiot-ic mycorrhizal fungi in a model ecosystem. Sci Rep 4:5634

Environ Sci Pollut Res