The rubromycins, such as gamma-rubromycin, heliquinomycin, and griseorhodin A, are a family of extensively modified aromatic polyketides that inhibit HIV reverse transcriptase and human telomerase. Telomerase inhibition crucially depends on the presence of a spiroketal moiety that is unique among aromatic polyketides. Biosynthetic incorporation of this pharmacophore into the rubromycins results in a dramatic distortion of the overall polyketide structure, but how this process is achieved by the cell has been obscure. To identify the enzymes involved in spiroketal construction, we generated 14 gene-deletion variants of the griseorhodin A biosynthetic gene cluster isolated from the tunicate-associated bacterium Streptomyces sp. JP95. Heterologous expression and metabolic analysis allowed for an assignment of most genes to various stages of griseorhodin tailoring and pharmacophore generation. The isolation of the novel advanced intermediate lenticulone, which exhibits cytotoxic, antibacterial, and elastase-inhibiting activity, provided direct evidence that the spiroketal is formed by cleavage of four carbon-carbon bonds in a pentangular polyketide precursor. This remarkable transformation is followed by an epoxidation catalyzed by an unusual cytochrome P450/NADPH:ubiquinone oxidoreductase pair that utilizes a saturated substrate. In addition, the absolute configuration of griseorhodin A was determined by quantum-chemical circular dichroism (CD) calculations in combination with experimental CD measurements.

The benastatins, pradimicins, fredericamycins, and members of the griseorhodin/rubromycin family represent a structurally and functionally diverse group of long-chain polyphenols from actinomycetes. Comparison of their biosynthetic gene clusters (ben, prm, fdm, grh, rub) revealed that all loci harbor genes coding for a similar, yet uncharacterized, type of ketoreductases. In a phylogenetic survey of representative KRs involved in type II PKS systems, we found that it is generally possible to deduce the KR regiospecificity (C-9, C-15, C17) from the amino acid sequence and thus to predict the nature of the aromatic polyketide (e.g., angucycline, anthracycline, benzoisochromanequinones). We hypothezised that the new clade of KRs is characteristic for biosynthesis of polyphenols with an extended angular architecture we termed "pentangular". To test this hypothesis, we demonstrated the biogenetic relationship between benastatin and the structurally unrelated spiro ketal griseorhodin by generating a mutant producing collinone, a pentangular pathway intermediate. The benastatin pathway served as a model to characterize the KR. Gene inactivation of benL resulted in the formation of a series of 19-hydroxy benastatin and bequinostatin derivatives (e.g., benastatin K and benastatin L). These results clearly showed that BenL functions as a C-19 KR in pentangular pathways.