Rethinking ‘secondary’ metabolism: physiological roles for phenazine antibioticsAlexa Price-Whelan1, Lars E P Dietrich2 & Dianne K Newman2,3Microorganisms exist in the environment as multicellular communities that face the challenge of surviving under nutrient-limited conditions. Chemical communication is an essential part of the way in which these populations coordinate their behavior, and there has been an explosion of understanding in recent years regarding how this is accomplished. Much less, however, is understood about the way these communities sustain their metabolism. Bacteria of the genus Pseudomonas are ubiquitous, and are distinguished by their production of colorful secondary metabolites called phenazines. In this article, we suggest that phenazines, which are produced under conditions of high cell density and nutrient limitation, may be important for the persistence of pseudomonads in the environment.Divisions of 1Biology and 2Geological and Planetary Sciences, and 3Howard Hughes Medical Institute, California Institute of Technology, Pasadena, California 91125, USA. Correspondence should be addressed to D.K.N. ([email protected]).Published online 18 January 2006; corrected 7 March 2006 (details online); doi:10.1038/nchembio764Historically, microbiologists and chemists alike have categorized as ‘secondary metabolites’ a broad class of molecules produced at late stages of microbial growth in laboratory cultures. This nomen-clature is, admittedly, pejorative, implying that these molecules are somehow less important than others to the cell that produces them. In particular, the traditional view is that secondary metabolites (i) do not contribute to the growth or survival of the producer (ii) are highly sensitive to the conditions stimulating their production (for example, medium composition) (iii) often have complex structures and (iv) have production rates that are decoupled from the doubling time of the cells1,2. Together, these leitmotifs present a conundrum: why would an organism limited for nutrients begin excreting large amounts of complex organic molecules? One reasonable and com-monly stated answer is that they function as antibiotics, and are pro-duced in copious quantities at this stage of growth to protect the producer from competitors3. In recent years, however, the idea that ‘secondary’ metabolites might have other functions, ranging from controlling gene expression4 to supporting growth or iron acquisition in microbial communities5,6, has become increasingly compelling. This is due, in large part, to the recognition that microbes typically exist in nature in biofilm communities7 and/or in a metabolically quiescent state8; because the ‘rules of the game’ for metabolism under these conditions are virtually unknown, a re-examination of the func-tion of secondary metabolites is warranted.To illustrate the idea that secondary metabolites have the potential to perform primary metabolic functions, we will focus this review on a class of compounds known as ‘phenazines’, which have been of great inter-est to pharmaceutical and clinical research groups for the last 50 years9. Phenazines are heterocyclic compounds that are produced naturally and substituted at different points around their rings by different bacte-rial species (Table 1). Small modifications of the core phenazine structure give rise to a full spectrum of colors, ranging from the deep red of 5-methyl-7-amino-1-carboxyphenazinium betaine (aeruginosin A, 1) to the lemon yellow of phenazine-1-carboxylic acid (PCA, 2), to the bright blue of 1-hydroxy-5-methylphenazine (pyocyanin, 3) (Fig. 1). The combination and variety of functional groups added also determine the redox potential and solubility of these compounds, thus affecting their biological activity9–11.The antagonistic effects of almost all of these derivatives are usually attributed to one general characteristic: redox activity. The 2-hydroxy-phenazine-1-carboxylic acid (2-OHPCA, 4) produced by Pseudomonas aureofaciens is thought to kill off competing fungi through the pro-duction of reactive oxygen species12. Many of the effects of pyocyanin (PYO) and PCA on a diversity of eukaryotic hosts as well as bacteria are thought to result from oxidative activity or the inactivation of proteins important in the oxidative stress response13,14. Regardless of whether they are acting as antibiotics in the soil, or virulence factors during infection, the redox transformations of phenazines strongly influence their physiological effects in other organisms. A more detailed understanding of phenazine metabolism in competing or host cells is emerging, as very recently, researchers have begun to recognize that small variations in the reactivity of phenazines can give rise to differences in their elicited response15.Concomitant with the development of ideas about phenazine activity during competition and infection, Pseudomonas aeruginosa and other phenazine-excreting bacteria have become popular model organisms for the study of quorum sensing and biofilm formation, two of the most active areas of research in the field of microbiol-ogy16–18. Whereas pharmaceutical and clinical groups have been focused on the physiological effects of these compounds in non-producing organisms, microbial physiologists and geneticists have typically viewed phenazines as metabolites that perform only second-ary functions. As a result, despite research interest in both the bio-NATURE CHEMICAL BIOLOGY VOLUME 2 NUMBER 2 FEBRUARY 2006 71REVIEW© 2006 Nature Publishing Group http://www.nature.com/naturechemicalbiologylogical activity of the compounds themselves and in the physiology of their producers, the primary functions of phenazines for produc-ing organisms such as the pseudomonads are still unknown. This is surprising, especially given that phenazine production and reduction is evident in many of the Pseudomonas cultures that microbiologists prepare for their work (Fig. 2) and that the mechanisms thought to be responsible for phenazine metabolism in non-producers (for example, reduction by NADH or glutathione, interaction with the respiratory chain) are present in most organisms19,20. That phen-azines and other excreted compounds (i) react with common pri-mary metabolites (ii) are potentially transformed by enzymes active in central metabolic pathways and (iii) induce gene expression calls into question their categorization as secondary metabolites. We