N-Acetylglucosamine, the monomer of chitin, is a favored carbon and nitrogen source for streptomycetes. Its intracellular catabolism requires the combined actions of the N-acetylglucosamine-6-phosphate (GlcNAc-6P) deacetylase NagA and the glucosamine-6-phosphate (GlcN-6P) deaminase/isomerase NagB. GlcNAc acts as a signaling molecule in the DasR-mediated nutrient sensing system, activating development and antibiotic production under poor growth conditions (famine) and blocking these processes under rich conditions (feast). In order to understand how a single nutrient can deliver opposite information according to the nutritional context, we carried out a mutational analysis of the nag metabolic genes nagA, nagB, and nagK. Here we show that the nag genes are part of the DasR regulon in Streptomyces coelicolor, which explains their transcriptional induction by GlcNAc. Most likely as the result of the intracellular accumulation of GlcN-6P, nagB deletion mutants fail to grow in the presence of GlcNAc. This toxicity can be alleviated by the additional deletion of nagA. We recently showed that in S. coelicolor, GlcNAc is internalized as GlcNAc-6P via the phosphoenolpyruvate-dependent sugar phosphotransferase system (PTS). Considering the relevance of GlcNAc for the control of antibiotic production, improved insight into GlcNAc metabolism in Streptomyces may provide new leads toward biotechnological applications.

Moonmilk is a karstic speleothem mainly composed of fine calcium carbonate crystals (CaCO3) with different textures ranging from pasty to hard, in which the contribution of biotic rock-building processes is presumed to involve indigenous microorganisms. The real microbial input in the genesis of moonmilk is difficult to assess leading to controversial hypotheses explaining the origins and the mechanisms (biotic vs. abiotic) involved. In this work, we undertook a comprehensive approach in order to assess the potential role of filamentous bacteria, particularly a collection of moonmilk-originating Streptomyces, in the genesis of this speleothem. Scanning electron microscopy (SEM) confirmed that indigenous filamentous bacteria could indeed participate in moonmilk development by serving as nucleation sites for CaCO3 deposition. The metabolic activities involved in CaCO3 transformation were furthermore assessed in vitro among the collection of moonmilk Streptomyces, which revealed that peptides/amino acids ammonification, and to a lesser extend ureolysis, could be privileged metabolic pathways participating in carbonate precipitation by increasing the pH of the bacterial environment. Additionally, in silico search for the genes involved in biomineralization processes including ureolysis, dissimilatory nitrate reduction to ammonia, active calcium ion transport, and reversible hydration of CO2 allowed to identify genetic predispositions for carbonate precipitation in Streptomyces. Finally, their biomineralization abilities were confirmed by environmental SEM, which allowed to visualize the formation of abundant mineral deposits under laboratory conditions. Overall, our study provides novel evidences that filamentous Actinobacteria could be key protagonists in the genesis of moonmilk through a wide spectrum of biomineralization processes.

Iron is one of the most abundant elements on earth but is found in poorly soluble forms hardly accessible to microorganisms. To subsist, they have developed iron-chelating molecules called siderophores that capture this element in the environment and the resulting complexes are internalized by specific uptake systems. While biosynthesis of siderophores in many bacteria is regulated by iron availability and oxidative stress, we describe here a new type of regulation of siderophore production. We show that in Streptomyces coelicolor, their production is also controlled by N-acetylglucosamine (GlcNAc) via the direct transcriptional repression of the iron utilization repressor dmdR1 by DasR, the GlcNAc utilization regulator. This regulatory nutrient-metal relationship is conserved among streptomycetes, which indicates that the link between GlcNAc utilization and iron uptake repression, however unsuspected, is the consequence of a successful evolutionary process. We describe here the molecular basis of a novel inhibitory mechanism of siderophore production that is independent of iron availability. We speculate that the regulatory connection between GlcNAc and siderophores might be associated with the competition for iron between streptomycetes and their fungal soil competitors, whose cell walls are built from the GlcNAc-containing polymer chitin. Alternatively, GlcNAc could emanate from streptomycetes' own peptidoglycan that goes through intense remodelling throughout their life cycle, thereby modulating the iron supply according to specific needs at different stages of their developmental programme.

Biosynthetic gene clusters (BGCs) are organized groups of genes involved in the production of specialized metabolites. Typically, one BGC is responsible for the production of one or several similar compounds with bioactivities that usually only vary in terms of strength and/or specificity. Here we show that the previously described ferroverdins and bagremycins, which are families of metabolites with different bioactivities, are produced from the same BGC, whereby the fate of the biosynthetic pathway depends on iron availability. Under conditions of iron depletion, the monomeric bagremycins are formed, representing amino-aromatic antibiotics resulting from the condensation of 3-amino-4-hydroxybenzoic acid with p-vinylphenol. Conversely, when iron is abundantly available, the biosynthetic pathway additionally produces a molecule based on p-vinylphenyl-3-nitroso-4-hydroxybenzoate, which complexes iron to form the trimeric ferroverdins that have anticholesterol activity. Thus, our work shows a unique exception to the concept that BGCs should only produce a single family of molecules with one type of bioactivity and that in fact different bioactive molecules may be produced depending on the environmental conditions. IMPORTANCE Access to whole-genome sequences has exposed the general incidence of the so-called cryptic biosynthetic gene clusters (BGCs), thereby renewing their interest for natural product discovery. As a consequence, genome mining is the often first approach implemented to assess the potential of a microorganism for producing novel bioactive metabolites. By revealing a new level of complexity of natural product biosynthesis, we further illustrate the difficulty of estimation of the panel of molecules associated with a BGC based on genomic information alone. Indeed, we found that the same gene cluster is responsible for the production of compounds which differ in terms of structure and bioactivity. The production of these different compounds responds to different environmental triggers, which suggests that multiplication of culture conditions is essential for revealing the entire panel of molecules made by a single BGC.

Actinobacteria are prolific producers of antitumor antibiotics with antiproliferative activity, but why these bacteria synthetize metabolites with this bioactivity has so far remained a mystery. In this work we raised the hypothesis that under certain circumstances, production of antiproliferative agents could be part of a genetically programmed death of the producing organism. While programmed cell death (PCD) has been well documented when Streptomyces species switch from vegetative (nutrition) to aerial (reproduction) growth, lethal determinants are yet to be discovered. Using DNA-damaging prodiginines of Streptomyces coelicolor as model system, we revealed that, under certain conditions, their biosynthesis is always triggered in the dying zone of the mycelial network prior to morphological differentiation, right after an initial round of cell death. The programmed massive death round of the vegetative mycelium is absent in a prodiginine non-producer (ΔredD strain), and mutant complementation restored both prodiginine production and cell death. The redD null mutant of S. coelicolor also showed increased DNA, RNA, and proteins synthesis when most of the mycelium of the wild-type strain was dead when prodiginines accumulated. Moreover, addition of the prodiginine synthesis inhibitors also resulted in enhanced accumulation of viable filaments. Overall, our data enable us to propose a model where the time-space production of prodiginines is programmed to be triggered by the perception of dead cells, and their biosynthesis further amplifies the PCD process. As prodiginine production coincides with the moment S. coelicolor undergoes morphogenesis, the production of these lethal compounds might be used to eradicate the obsolete part of the population in order to provide nutrients for development of the survivors. Hence, next to weapons in competition between organisms or signals in inter- and intra-species communications, we propose a third role for antibiotics (in the literal meaning of the word ‘against life’) i.e., elements involved in self-toxicity in order to control cell proliferation, and/or for PCD associated with developmental processes.