In this work, we have generated a first proteomic map of Pseudozyma flocculosa, a biocontrol fungus known for its antagonistic activity against powdery mildew fungi mediated, at least in part, by the release of an antifungal glycolipid, flocculosin. Our aim was to compare the proteomic profile of cell extracts grown under two specific morphological conditions, including one conducive to flocculosin synthesis, in order to gain a better understanding of the factors that could trigger the biocontrol activity of P. flocculosa.
Our results highlighted important qualitative and quantitative changes in the synthesis of individual proteins during the fungus' adaptation to a carbon-supplemented medium leading to flocculosin synthesis. As a first observation, we noted a clear decrease in the total number of proteins in the stressed cells, compared to control cells. This is indicative of an important reduction in certain metabolic activities that would be linked with the release of stress metabolites, namely flocculosin. Incidentally, Mimee et al.
 have recently proposed that flocculosin could serve both roles of niche protection and food reserve under limiting conditions. This difference in metabolic behaviour was also accompanied with a gradual transition from the yeast-like to the filamentous form (see Fig. 1), the latter only being apt to produce flocculosin. While this transition step is not accompanied by an increase in biomass, as would be expected in typical filamentous fungi such as molds, Hammami et al.
 suggested that it was induced as an adaptation/protection stage to a new environment for the epiphytic fungus. In line with these observations, some proteins associated with flocculosin biosynthesis in P. flocculosa were detected specifically in cells producing the glycolipid. For instance, a septin-like protein was recorded exclusively in cells grown in the carbon amended medium. Septins are GTPases that form filaments in fungi and animals, and are thought to function in controlling cytokinesis and coordinating nuclear division and membrane movement [11, 12]. Together with septin, we also detected a Ras-like protein reported to promote filamentous growth in U. maydis
Given that flocculosin synthesis was supported by C-fed-batch, it is also normal that several protein activities found in this treatment were linked to carbon metabolism. The expression levels of transketolase and transaldolase, for instance were 2 and 2.5 fold higher, respectively in the flocculosin-inducing medium. These two enzymes are involved in the non-oxidative part of the pentose phosphate pathway. The pentose phosphate pathway involves two steps: the first one (oxidative part) metabolizes glucose-6-phosphate to ribulose-5-phosphate, the principal carbon source for nucleotide synthesis and NADPH for reductive reactions in biosynthesis and perhaps for hydroxylation . When the cell is not in suitable conditions to produce biomass, ribulose-5-phosphate joins the second step of the pathway (non-oxidative part), to be converted to glyceraldehyde-3-phosphate and two fructose-6-phosphate. The detection of thiamine biosynthesis activity in the flocculosin inducing medium is consistent with an increase in transketolase levels. Thiamine contributes to carbon metabolism by acting as cofactor of several enzymes such as pyruvate dehydrogenase, α-ketoglutarate dehydrogenase, branched-chain α-ketoacid dehydrogenase, transketolase and pyruvate decarboxylase . Based on this, we can speculate that P. flocculosa does undergo a metabolic switch from biomass production to flocculosin production under stress conditions as proposed by Hammami et al.
. Indeed, the latter study found that the presence of yeast extract supported the conidial biomass production of P. flocculosa. When growth is blocked by the exhaustion of yeast extract, flocculosin excretion might constitute an overflow metabolism for P. flocculosa that regulates the intracellular energy level.
In addition to the observation of proteins consistent with the inhibition of biomass production under stress conditions, several other proteins were induced or increased in P. flocculosa cells following the addition of a carbon source in the medium. Higher levels of glutamate synthase, cyanate lyase and heat shock protein (HSP) 70 and 90 support the hypothesis that P. flocculosa was under stress conditions following the fed-batch. The addition of carbon increases the consumption speed of the components of the complex medium YMPD including the yeast extract ingredients. The exhaustion of these components, particularly yeast extract, appears to induce flocculosin synthesis, a strategy that may have been developed by P. flocculosa to store excess carbon in the environment as proposed by Mimee et al.
. As explained earlier, the carbon backbone of flocculosin (and of ustilagic acid) is composed of a long fatty acid chain originating from palmitic acid. Here, we detected the electron transfer flavoprotein, a key cofactor of the β-oxidation which is considered a way to produce acetyl-COA in the mitochondria. Equally, the expression of pentose phosphate-related enzymes is necessary to produce NADPH, which in turn is essential to fatty acid synthesis. Because of the complexity of the flocculosin (glycolipid) structure, its synthesis requires ATP production. Expectedly, an ATPase and other proteins such as the glycolytic enzyme, triose phosphate isomerase (TIM), succinate dehydrogenase and alpha-glucosidase involved in carbon metabolism producing the ATP molecule, were up-regulated by the carbon addition.
In the control medium, the cells tended to produce proteins that are commonly associated with biomass production. Ornithine aminotransferase has a fundamental role in the central metabolism of organisms, as it may serve for the synthesis of proline, polyamines, glutamate and glutamine. These products are of great nutritional and physiological importance for diverse functions including growth and development . Also, the soluble elongation factors are likely to play an important role in limiting the error frequency during protein synthesis . The low abundance of these proteins in the stress treatment suggests that carbon addition in the growth medium inhibited some activities which assure the appropriate course of cellular growth and shifted the overall metabolism toward the production of flocculosin.
Owing to the structural similarity between flocculosin and ustilagic acid, and to the phylogenetic link between P. flocculosa and U. maydis, we expected to detect in the flocculosin production treatment proteins belonging to the cluster of 12 co-regulated genes involved in ustilagic acid synthesis in U. maydis
. Our analyses did not identify any of the proteins related to this cluster of genes, whether they are present in the cytosol or in the membranes. The detection of membrane proteins such as the ATP-synthase seem to rule out the possibility that our extraction procedure failed to detect such proteins . Their low relative concentration or a poor correlation between mRNA transcripts and protein accumulation for these 12 genes, as noted for many genes in yeast , could explain our results. A similar conclusion was reported in the study of the proteomic changes involved in the transition of U. maydis from budding to filamentous form . Two proteins, versicolorin B synthase (UM03246) and anthranilate synthase (UM02376) that were minor components of multiprotein spots showed an increase in transcriptional level of 4- and 13-fold respectively.
Based on our observations, it does appear that P. flocculosa will only produce flocculosin early in the process of pseudohyphal formation under conditions of nutrient limitations. While this process can be artificially maintained in vitro by the addition of a C source, in nature it would serve as a means of protecting/forging an ecological niche as sporidia land in a new environment. This supports recent quantitative evaluation of flocculosin production in situ where Marchand et al.
 found that genes involved in flocculosin synthesis were only up-regulated within the first 6 hours of sporidia coming into contact with a leaf. In the same manner, we can speculate that sporidia of U. maydis would produce ustilagic acid in the first steps of germination to facilitate contact with a compatible mating type, as suggested by Hewald et al.