Search

Cultivated bacteria such as actinomycetes are a highly useful source of biomedically important natural products. However, such 'talented' producers represent only a minute fraction of the entire, mostly uncultivated, prokaryotic diversity. The uncultured majority is generally perceived as a large, untapped resource of new drug candidates, but so far it is unknown whether taxa containing talented bacteria indeed exist. Here we report the single-cell- and metagenomics-based discovery of such producers. Two phylotypes of the candidate genus 'Entotheonella' with genomes of greater than 9 megabases and multiple, distinct biosynthetic gene clusters co-inhabit the chemically and microbially rich marine sponge Theonella swinhoei. Almost all bioactive polyketides and peptides known from this animal were attributed to a single phylotype. 'Entotheonella' spp. are widely distributed in sponges and belong to an environmental taxon proposed here as candidate phylum 'Tectomicrobia'. The pronounced bioactivities and chemical uniqueness of 'Entotheonella' compounds provide significant opportunities for ecological studies and drug discovery.

Nature 506, 58–62 (2014); doi:10.1038/nature12959 One of the accession numbers for this Article was listed as AZHXW01000000 instead of AZHX01000000. It has been corrected in the online versions of the paper.

Marine sponges are a rich source of bioactive natural products and are promising sources for drug discovery and development An impressive example is the sponge Theonella swinhoei, which has yielded more than 120 compounds belonging to diverse structural types. Many sponges also harbor highly complex consortia of symbiotic bacteria that are suspected to be the true source of at least some of the secondary metabolites. In previous work, our group demonstrated a bacterial origin of onnamide- and psymberin-type polyketides for two different sponges (1–3), but there were no insights into the producer of compounds from other natural product families. In addition, the exact taxonomic identity of sponge-associated producers remained unknown. This talk will present new insights into these two issues. Isolation of genes encoding a peptide biosynthetic pathway from the T. swinhoei metagenome demonstrated a bacterial origin. Several genes were heterologously expressed and functionally characterized, which revealed unprecedented biosynthetic transformations. The novelty of these modifications suggests the existence of a structurally distinct natural product family, for which we propose the name proteusins. Using a strategy consisting of single-cell analysis and metagenomic sequencing, we identified the bacterial producer of onnamide polyketides in T. swinhoei. Surprisingly, the data suggest the symbiont to be a chemically exceptionally prolific bacterium, producing not only onnamides but most other compounds from this sponge chemotype, including the known and two previously unknown proteusins. Further biosynthetic studies and a survey of other sponges indicate that close relatives of the producer are widespread in these animals and vary with respect to their biosynthetic capabilities. These bacteria might therefore represent the first uncultivated taxon with a metabolic richness resembling that of major cultivated bacterial natural product sources. These results reveal a key role of symbiotic bacteria in the chemistry of their sponge hosts and provide new strategies to study uncultivated symbionts in a more systematic fashion.

Many marine sponges harbor massive consortia of symbiotic bacteria belonging to diverse phyla. Sponges are also an unusually rich source of biologically active natural products, and evidence is accumulating that these compounds might often be synthesized by the symbionts. Since the study of sponge-associated bacteria is generally hampered by very low cultivation rates, cultivation-independent, metagenomic methods have recently been applied to sponges. These methods allow for the isolation of biosynthetic gene clusters that can ultimately be exploited to develop sustainable natural product sources by heterologous expression. However, general challenges encountered in sponge metagenomic research are the poor quality of the isolated DNA with respect to size and yield, the difficulty to identify genes of interest among numerous homologs, insufficient clone numbers in metagenomic libraries, and time-consuming screening procedures to identify and isolate rare positive clones. Here, we give an overview of methods that address these problems and can be used to streamline isolation of biosynthetic and other genes of interest.

Corallopyronin A is a myxobacterial compound with potent antibacterial activity. Feeding experiments with labelled precursors resulted in the deduction of all biosynthetic building blocks for corallopyronin A and revealed an unusual feature of this metabolite: its biosynthesis from two chains, one solely PKS-derived and the other NRPS/PKS-derived. The starter molecule is believed to be carbonic acid or its monomethyl ester. The putative corallopyronin A biosynthetic gene cluster is a trans-AT-type mixed PKS/NRPS gene cluster, containing a beta-branching cassette. Striking features of this gene cluster are a NRPS-like adenylation domain that is part of a PKS-type module and is believed to be responsible for glycine incorporation, as well as split modules with individual domains occurring on different genes. It is suggested that CorB is a trans-acting ketosynthase and it is proposed that it catalyses the Claisen condensation responsible for the interconnection of the two chains. Additionally, the stereochemistry of corallopyronin A was deduced by a combination of a modified Mosher's method and ozonolysis with subsequent chiral GC analyses.

Many marine sponges harbor massive consortia of symbiotic bacteria belonging to diverse phyla. Sponges are also an unusually rich source of biologically active natural products, and evidence is accumulating that these compounds might often be synthesized by the symbionts. Since the study of sponge-associated bacteria is generally hampered by very low cultivation rates, cultivation-independent, metagenomic methods have recently been applied to sponges. These methods allow for the isolation of biosynthetic gene clusters that can ultimately be exploited to develop sustainable natural product sources by heterologous expression. However, general challenges encountered in sponge metagenomic research are the poor quality of the isolated DNA with respect to size and yield, the difficulty to identify genes of interest among numerous homologs, insufficient clone numbers in metagenomic libraries, and time-consuming screening procedures to identify and isolate rare positive clones. Here, we give an overview of methods that address these problems and can be used to streamline isolation of biosynthetic and other genes of interest.

The pharmacological effects of leaf extracts (EGb 761) from GINKGO BILOBA L. are attributed to ginkgolides, bilobalide and biflavonoids. However, besides these beneficial attributes, ginkgotoxin, a B 6 antivitamin which may cause epileptic convulsions, other severe neuronal disorders and even death, is also found in GINKGO leaves and leaf-derived remedies. Because of its structural similarity to the B 6 vitamers, an interaction of ginkgotoxin with enzymes involved in the vitamin B 6 -dependent metabolism of the human brain is possible. This led us to investigate how the neurotoxic ginkgotoxin acts in the brain. To this end the gene coding for the human pyridoxine 5′-phosphate oxidase was heterologously overexpressed in E. COLI and the homogeneous enzyme was characterized. The investigation showed that the enzyme is inhibited IN VITRO by the synthetic vitamin B 6 derivative 4′-deoxypyridoxine 5′-phosphate but not by ginkgotoxin or its 5′-phosphate.

Vitamin B6 represents a highly important group of compounds ubiquitous in all living organisms. It has been demonstrated to alleviate oxidative stress and in its phosphorylated form participates as a cofactor in >100 biochemical reactions. By means of a genetic approach, we have identified a novel mutant, rsr4-1 (for reduced sugar response), with aberrant root and leaf growth that requires supplementation of vitamin B6 for normal development. Cloning of the mutated gene revealed that rsr4-1 carries a point mutation in a member of the PDX1/SOR1/SNZ (for Pyridoxine biosynthesis protein 1/Singlet oxygen resistant 1/Snooze) family that leads to reduced vitamin B6 content. Consequently, metabolism is broadly altered, mainly affecting amino acid, raffinose, and shikimate contents and trichloroacetic acid cycle constituents. Yeast two-hybrid and pull-down analyses showed that Arabidopsis thaliana PDX1 proteins can form oligomers. Interestingly, the mutant form of PDX1 has severely reduced capability to oligomerize, potentially suggesting that oligomerization is important for function. In summary, our results demonstrate the critical function of the PDX1 protein family for metabolism, whole-plant development, and vitamin B6 biosynthesis in higher plants.

We studied the metabolism of radioactively labeled safingol (l-threo-dihydrosphingosine) in primary cultured neurons, B104 neuroblastoma cells, and Swiss 3T3 fibroblasts, and compared it to that of its natural stereoisomer d-erythro-dihydrosphingosine. Both sphingoid bases are used as biosynthetic precursors for complex sphingolipids, albeit to different rates. Whereas a considerable amount of the natural sphingoid base is also directed to the catabolic pathway (20-66%, cell type dependent), only a minor amount of the nonnatural safingol is subjected to catabolic cleavage, most of it being N-acylated to the respective stereochemical variant of dihydroceramide. Interestingly, N-acylation of safingol to l-threo-dihydroceramide is less sensitive to fumonisin B1 than the formation of the natural d-erythro-dihydroceramide. In addition, safingol-derived l-threo-dihydroceramide, unlike its physiologic counterpart, is not desaturated. Most of it either accumulates in the cells (up to 50%) or is used as a biosynthetic precursor of the respective dihydrosphingomyelin (up to 45%). About 5% is, however, glucosylated and channeled into the glycosphingolipid biosynthetic pathway. Our results demonstrate that, despite its nonnatural stereochemistry, safingol is recognized and metabolized preferentially by enzymes of the sphingolipid biosynthetic pathway. Furthermore, our data suggest that the cytotoxic potential of safingol is reduced rather than enhanced via its metabolic conversion.