Introduction
Dictyostelium discoideum1 is a free-living amoeba during its vegetative growth phase, but upon starvation it switches to a multicellular developmental phase that is induced by cyclic AMP (http://www.dictybase.org). The multicellular aggregate forms a slug that migrates toward heat and light. It then develops into a fruiting body made up of different cell types that form a mass of spore cells supported atop a stalk. Small signaling molecules called differentiation-inducing factors (DIFs) orchestrate the differentiation of cell types, such as the prespore and prestalk cells, in the multicellular slug2. DIF-1 (Fig. 1a) is a well-characterized polyketide-based molecule that induces the differentiation of pstO, one type of prestalk cell3. Taking advantage of the recently available genome sequence of D. discoideum, Austin et al. have identified the enzyme (named "Steely2") that is responsible for the synthesis of the DIF scaffold4.
Figure 1: DIF-1 synthesis in D. discoideum.
(a) The type III PKS of Steely2 synthesizes the PCP scaffold, which is further modified by chlorination and methylation reactions to form DIF-1. (b) The domain organization of Steely2, in which typical FAS domains (KS, ketosynthase; AT, acyltransferase; DH, dehydratase; ER, enoylreductase; KR, ketoreductase and ACP, acyl carrier protein) are associated with a type III PKS module.
Full size image (21 KB)Polyketides are chemically and structurally diverse secondary metabolites that are synthesized by polyketide synthases (PKSs) present in both prokaryotes and eukaryotes5. Recently, attention has been focused on both the ability of PKSs to synthesize complex chemical scaffolds and the possibility of engineering them to produce new antimicrobial compounds6. PKSs generate diverse molecules by taking advantage of reactivities in a way similar to that of fatty acid synthases (FASs), which carry out an iterative cycle of two-carbon chain extension through a ketoacyl synthase domain. However, compared to FASs, PKSs have an expanded capacity because they are able to off-load products after partial processing without going through a complete set of reactions, which can include steps such as reductions and dehydrations. Type III PKSs are small homodimeric enzymes that perform multiple enzymatic activities at a single site7. Chalcone synthase (CHS) from the plant Medicago sativa, which performs the first step in flavonoid biosynthesis, is one biochemically and structurally well-characterized type III system. Type III PKSs that belong to the CHS family were initially thought to be specific to plants, in which they produce the chalcone precursor for a variety of molecules involved in diverse roles such as microbial defense, UV protection and pigment formation. However, with the availability of sequenced genomes of many organisms, it is becoming clear that type III PKSs are present in several bacteria and other organisms as well, and some of them produce functionally important secondary metabolites. For example, RppA, the first type III PKS to be identified and characterized from a bacterial source (Streptomyces griseus), is involved in melanin production. A more recent report of the functional role of a type III PKS identified outside the plant kingdom is in the soil bacterium Azotobacter vinelandii, in which the enzyme has a role in formation of the cyst, a protective lipid coat8.
DIF-1 is based on a phloroglucinol skeleton and has been presumed to arise by a pathway involving PKS synthesis of phlorocaprophenone (PCP; Fig. 1a). However, until now few molecular details have been elucidated for the biosynthetic pathway that produces DIF-1, other than for the enzyme responsible for the final methylation step. Sequencing of the D. discoideum genome revealed the presence of nearly 40 genes homologous to PKSs. Using bioinformatic tools, Austin et al. identified two open reading frames, which they named "Steely1" and "Steely2," in which a type III PKS was associated with typical type I FAS domains. Surprisingly, in Steely2, the type III PKS module at its C terminus replaces the thioesterase domain normally expected at that position in most modular enzymes (Fig. 1b). The authors hypothesized that the FAS domains could provide a hexanoyl coenzyme A (CoA) starter molecule to the type III PKS, which could then carry out three cycles of acetyl group extension and cyclization to produce the phloroglucinol scaffold necessary for DIF synthesis. To test this, they performed in vitro assays in which hexanoyl CoA–primed reactions with Steely2 produced the DIF scaffold, but those with Steely1 did not. To further show that one of these proteins is actually responsible for DIF scaffold synthesis, they created mutant strains in which the Steely proteins were individually disrupted. Whereas the Steely1- mutant was not defective in DIF-1 synthesis, the Steely2- mutants did not produce DIF-1 and resulted in a developmental phenotype expected for cells lacking DIF-1. Supplementing Steely2- mutants with DIF-1 restored normal development, which supports the assignment of Steely2 as being responsible for the synthesis of DIF-1. The crystal structure of the type III PKS domain of Steely1 confirmed that it has the characteristic fold and the catalytic triad as observed in the related CHS family of enzymes.
Although the work identifies the enzyme for DIF-1 precursor synthesis, it seems as though it is the tip of the iceberg. DIF-1 is responsible for the differentiation of only one out of many D. discoideum cell types, and more DIFs are being identified from this organism9. It is likely that these chemically similar molecules are produced by similar pathways, yet it remains to be seen exactly how they are produced and regulated for the differentiation of multiple cell types. Therefore, this study by Austin et al. provides a starting point for probing the biochemical and functional role of the putative PKS genes, including that of Steely1, and others responsible for DIF production in D. discoideum.
The metabolite diversity generated by subtle changes in some of the type III PKSs that have been characterized in vitro, for example from Mycobacterium tuberculosis10, highlights the limitations of homology modeling and even of experimentally determined structures (in some cases) in predicting the diversity generated by PKS enzymes. Therefore, it is imperative to probe these enzymes using a combination of in vivo and in vitro techniques, such as was used by Austin et al., to obtain a holistic view of their function. It is also worth noting that domain swapping in modular PKSs has been used as a strategy to generate new polyketides6. The newly identified role of type III PKSs in a modular system in which they replace the thioesterase domain to allow release of growing polyketide chains adds another dimension to the metabolite diversity that can be generated by domain-swapping experiments. Such a methodology would also benefit from the emerging structural information on some of the modular FAS systems. Thus, the natural occurrence of these domains in varied contexts further highlights the potential of using the domain-swapping strategy for generating chemical diversity.
