Foliar hyperparasites of rust fungi and the genome mining of novel secondary metabolites

  • Gregory F. Harm

Western Sydney University thesis: Doctoral thesis

Abstract

Rust diseases caused by Puccinia spp. on cereal crops, such as wheat, can cause catastrophic crop losses across the globe, and have been identified as a threat to global food security. Given the prevalence, evolution and impact of rust diseases, there is a push to identify additional control methods. Hyperparasites (parasites of parasites) are a potential biological control method for rust fungi. Fungal hyperparasites also produce a variety of novel secondary metabolites (SMs), some of which are potent antifungals. SMs are produced by biosynthetic gene clusters (BGCs), and these clusters can be identified via genome mining. There are minimal genomes currently available from hyperparasites of rust fungi. There is also a lack of molecular information about fungal hyperparasites of rust fungi. Given the potential value of newly identified hyperparasites and/or novel SMs, molecular and genetic studies are required to address this gap. The key aims of this thesis were to: identify hyperparasites of rust, identify antifungal activity in putative hyperparasite crude extracts, isolate and characterise any antifungal compounds, sequence selected putative hyperparasites, and to mine the resultant genomes for BGCs. To address these aims, putative hyperparasites were isolated from the pustules of rust-infected plants: barley (infected with Puccinia hordei), oats (infected with Puccinia coronata f. sp. avenae), wheat (infected with Puccinia triticina, Puccinia striiformis f. sp. tritici, and Puccinia graminis f. sp. tritici)), willow (infected with Melamspora epitea), poplar (infected with Melampsora larici-populina), and myrtle (infected with Austropuccinia psidii). A total of 28 putative hyperparasites were isolated from rust infected plants, and 21 fungi were identified by sequencing the internal transcribed spacer (ITS) region. Eight isolates were selected for further analysis based on the absence of known phytopathogenicity. The selected eight isolates were additionally identified by sequencing their partial 18S ribosomal RNA and by morphological analysis. These were identified as: Clonostachys rosea (two isolates: HWLR12 and HOCR18), Lecanicillium psalliotae (HWLR35), Epicoccum nigrum (HPOP30 and HPOP33), Simplicillium aogashimaense (HWYR21), Neoascochyta sp. (HWLR27*), and Penicillium brevicompactum (HOCR17). The same eight putative hyperparasites were tested for antifungal activity against Fusarium equiseti, Aspergillus niger, Penicillium chrysogenum and Rhizopus sp. and analysed by liquid chromatography-mass spectroscopy (LC-MS). All crude extracts tested had varying antifungal activity, except for the S. aogashimaense HWYR21 extract, which had no antifungal activity. The compound oosporein was purified from the L. psalliotae HWLR35 isolate, and compounds 43_2A (a mannose moiety attached to a saturated fatty acid chain with an undetermined end group), and 43_3A (a phenyl containing polyketide) were purified from Neoascochyta sp. HWLR27*. Structural elucidation of 43_2A and 43_3A was not completed. A P. triticina urediniospore germination inhibition assay was used to determine the antifungal activity of the isolated compounds against rust fungi. In the germination inhibition assay, oosporein was found to qualitatively reduce the length of P. triticina urediniospore germ tubes, 43_3A was found to be minimally active, and 43_2A was found to have inhibition equal to the nystatin control. Both oosporein and 43_2A had antifungal activity against F. equiseti and P. chrysogenum at the same concentrations. Six of the eight fungal hyperparasites were genome sequenced: S. aogashimaense HWYR21 (29.7 Mb), Neoascochyta sp. HWLR27* (29.2 Mb), L. psalliotae HWLR35 (36.1 Mb), C. rosea HWLR12 (55.4 Mb), C. rosea HOCR18 (55.6 Mb), and Penicillium brevicompactum HOCR17 (30.5 Mb). HWYR21, HWLR27* and HWLR35 had not previously been sequenced. All of the genomes had more than 99% gene completeness by Benchmarking Universal Single-Copy Orthologs (BUSCOs). Genome mining for all classes of BGCs was performed using antiSMASH. A total of 579 predicted BGCs were identified, with 226 of those belonging to the main BGC classes of NRPS, PKS, and terpene synthases. Only three of the identified clusters had a complete match to characterised clusters in the Minimum Information about a Biosynthetic Gene cluster (MIBiG) database: the mycophenolic acid BGC in P. brevicompactum HOCR17, the oosporein BGC in L. psalliotae HWLR35, and the T-toxin BGC in Neoascochyta sp. HWLR27*. This is the first study to investigate multiple hyperparasites of rust to this level of genomic detail and demonstrates that these hyperparasites harbour a plethora of unknown and novel BGCs. The future expression and characterisation of the novel BGCs identified is essential as the SM products are potentially novel. This could lead to the identification of potent antifungal compounds for use in agriculture. This thesis also creates a starting point for future investigations into the metabolome and the transcriptome of these hyperparasites. The transcriptome could be used to evaluate the mechanisms deployed by these fungi during hyperparasitism and assist in the understanding of fungal biocontrol. Overall, this thesis highlights the value of discovery and characterisation of new fungal hyperparasites as a source of novel bioactive compounds.
Date of Award2019
Original languageEnglish

Keywords

  • cereal rusts
  • rust fungi
  • control
  • metabolites
  • antifungal agents

Cite this

'