Analysis of the genetic potential of the spongederived fungus <em>Penicillium chrysogenum</em> E01-10/3 for polyketide production

Abstract

The main goal of the presented dissertation was to evaluate the genetic potential of P. chrysogenum E01-10/3 strain for the production of polyketides. <br /> This marine-derived P. chrysogenum strain was isolated from the Mediterranean sponge Ircinia fasciculata in the course of a research program focused on the discovery and characterization of novel natural products. This led to the isolation and characterisation of two novel and structurally highly similar polyketides: sorbicillactone A and sorbicillactone B. Consistent with their structural similarity, it was proposed that the biosynthesis of both compounds might be encoded by the same gene cluster. Consequently, the identification of the sorbicillactone A gene cluster was in the focus of this dissertation as this compound was previously shown to posses promising antileukaemic, antiviral and neuroprotective properties. <br /> The iterative type I PKSs, which have only one multidomain protein with all the enzyme activities covalently bound together, are responsible for the biosynthesis of fungal polyketides. The single multifunctional protein is used to iteratively catalyze multiple rounds of chain elongation and appropriate β-keto processing of a polyketide.<br /> Since all of the fungal PKSs belong to the iterative type I PKS enzymes, degenerate primers and hybridization probes fitting to this type of fungal PKS systems were used in PCR and hybridization experiments. Of special help for the present study was the possibility to differentiate between subtypes of fungal iterative type I PKSs on the amino acid level: Nonreducing (NR), partially reducing (PR), and highly reducing (HR), in respect to level of reduction of their polyketide products. Accordingly, PCR and hybridization experiments were set up in order to take advantage of this fact.<br /> During the course of this study, use of PCR enabled the amplification of partial PKS sequences from nine putatively distinct fungal type I PKS gene loci from P. chrysogenum E01 10/3. Six partial KS domain DNA sequences were used to reconstruct evolutionary relationships in respect to other iterative type I PKSs. The results of the phylogenetic analysis for KS domains illustrated that P. chrysogenum E01-10/3 strain has the genetic potential to produce all three main categories of fungal polyketides – HR, PR (i.e. 6-MSA type) and NR. <br /> The results of the phylogenetic analysis of PCR-amplified partial KS domains were valuable to judge on good candidates for the screening of genomic library for putative sorbicillactone gene cluster: Three partial PKS sequences shown to be putative members of the NR clade III (KHKS1, KHKS32 and AT-9-11). For each of these three sequences specific primer pairs were designed and used in a complex genomic library screening procedure.<br /> The subcloning and sequencing of the AT-9-11 genomic region revealed that one putative reducing PKS and one non-reducing PKS were located next to each other. The genomic insert anticipated to carry the core of the putative AT-9-11 PKS gene cluster was completely sequenced and analysed.<br /> The sequence analysis of the putative sorbicillactone gene cluster identified eight putative ORFs matching fungal protein sequences from public databases. The first identified gene encodes for a putative transcriptional regulator (slr or orf1), which might be capable to coordinate expression of the structural genes in the cluster. This putative regulatory gene is followed by the gene encoding for one putative monooxygenase (slmox or orf2) that may be responsible for one post-PKS hydroxylation reaction during the sorbicillactones biosynthesis. The core of the identified gene cluster contains two PKS genes (PKSSL1/orf3 and PKSSL2/orf4) located next to each other in opposite directions. The second putative transcriptional regulatory gene (sltr or orf5) appears as a fifth ORF and is placed next to a putative gene for a MFS transporter protein (slMFS or orf6). The product encoded from the seventh ORF (slox or orf7) could be an oxidoreductase. The last identified ORF that showed putative homology to fungal proteins was orf8 that shared significant similarity with fungal actin cytoskeleton-regulatory complex proteins. This ORF was excluded from the predicted sorbicillactone gene cluster model, since its involvement in cytoskeleton dynamics is not likely to be required for the biosynthesis nor transport of sorbicillactones.<br /> The analysis of the exact domain organisation of the identified P. chrysogenum PKS genes – pksSL1 and pksSL2, enabled the prediction of their putative routes in the sorbicillactone A and B biosynthesis. The performed protein sequence analysis showed that the domain organisation of PKSsl1 is consistent with the one expected for the members of NR clade III: SAT-KS-AT-PT-ACP-MT-RED. All known PKS active site motifs were conserved and the analysed domains were not fragmented. These facts supported the prediction that they are active in P. chrysogenum PKSsl1. Additionally, the phylogenetic analysis of the PT domain gave clear hint that PKSsl1 could work as a tetraketide synthase.<br /> Based on the domain analysis it was confirmed that PKSsl2 has the typical domain structure of a R clade I PKS: KS-AT-DH-(MT)-ER-KR-ACP. Thus, PKSsl2 could be sufficient for biosynthesis of the advanced triketide starter unit, since it contains all domains required for condensation and the complete processing of a ß-carbon. All PKSsl2 domains have conserved consensus sequences within active sites and are not fragmented. However, there is a high chance that the detected PKSsl2 MT domain has no activity since the consensus sequence within its active site is not preserved. Such nonmethylated, reduced triketide starter unit could be produced by the PKSsl2 within two successive condensation and ß-processing rounds. This advanced starter unit may be then accepted by the SAT domain of PKSsl1 that probably makes three additional nonreducing extension rounds. At the end of the discussed putative sorbicillactone polyketide routes, it was proposed that hexaketide thiolester intermediates of sorbicillactone biosynthesis are released in an aldehyde form by the RED domain encoded reductase followed by the release of the free holo-ACP thiol of PKSsl1.<br /> Taking into account the number and potential bioactivities of polyketides isolated from marine-derived fungi, the number of compounds reported up to date from different P. chrysogenum strains as well as the genetic potential of P. chrysogenum E01-10/3 for polyketide production identified in this study, it becomes apparent that this particular marine-derived strain may be a valuable source for pharmacologically useful polyketides.