Björn Sandrock

Mannosylerythritol lipids (MELs) are antibacterial surface-active molecules produced by several basidiomycetous fungi. Enzymes for MEL biosynthesis are encoded in a gene cluster composed of five genes, which are both necessary and sufficient for the biosynthesis of MELs. Here, we report that in two closely related biotrophic pathogens —Sporisorium reilianum f. sp. reilianum ( SrS ) and Sporisorium reilianum SRZ2 —MEL production is achieved through the combined activity of two gene clusters. While the larger cluster is required for production of the disaccharide moiety, for attachment of one acyl chain to the sugar backbone and for export of MELs, the smaller cluster encodes two enzymes for additional acylation of the sugar backbone and is essential to produce mature MELs in S. reilianum . The genes in the smaller cluster appear to be duplications of two of the original genes in the larger cluster, which are not expressed in SrS but in related fungi, e.g., Ustilago maydis . MELs made by SrS are tri-acylated and exhibit altered physical and chemical properties, making them interesting candidates for future biotechnological and pharmaceutical applications. In summary, our study showcases how duplication and diversification of MEL clusters can lead to the evolution of novel biosurfactants. IMPORTANCE: Secondary metabolites produced by fungi can act as weapons against competitors, can help access nutrients, or assist development and communication. One group of secondary metabolites are surface-active glycolipids that have a great potential as biodegradable detergents. Upon nitrogen starvation, biosynthesis of the glycolipid mannosylerythritol lipid (MEL) is induced by a gene cluster composed of five genes in the plant pathogen Ustilago maydis and related basidiomycetous fungi. This study shows that in the smut fungus Sporisorium reilianum f. sp. reilianum, the five genes of the MEL gene cluster were not sufficient to produce the isolated MEL variant. In contrast to conventional MELs, MELs of S. reilianum are tri-acylated at the mannose moiety with acyl groups in the range of C ₆ –C ₁₀ . These MELs exhibit altered physical and chemical properties, making them interesting novel candidates for future applications. Furthermore, we demonstrate that in S. reilianum, a combination of two gene clusters is necessary for MEL biosynthesis, which enables a glimpse into the evolutionary history of the altered MEL species.

Peroxisomes are single membrane-bounded oxidative organelles with various metabolic functions including β-oxidation of fatty acids. Peroxisomes of many species confine certain metabolic enzymes into sub-compartments sometimes visible as electron dense cores. Why these structures form is largely unknown. Here, we report that in the smut fungus Ustilago maydis detergent resistant core structures are enriched for different enzymes excluding several key enzymes of the β-oxidation pathway. This confinement contributes to generation of peroxisome subpopulations that differ in their enzyme content. We identify short amino acid motifs necessary and sufficient for protein self-assembly into aggregates in vitro. The motifs trigger enrichment in cores in vivo and are active in mammalian cells. Perturbation of core assembly via variation of such motifs affects peroxisome function in U. maydis strains challenged with fatty acids. Thus, protein core structures serve to compartmentalize the lumen of peroxisomes thereby preventing interference of biochemical reactions. Metabolic compartmentalization of peroxisomes via assembly of specific proteins may occur in other organisms as well.

Fungi produce and secrete various secondary metabolites that can act as weapons against competitors, help in accessing nutrients, or assist in development and communication. One group of secondary metabolites are surface-active glycolipids, which have significant biotechnological potential as biodegradable detergents.

Formation of specialized reaction spaces prevents interference between distinct cellular pathways. Peroxisomes are cellular compartments involved in a large diversity of metabolic processes. How peroxisomes differentiate into subpopulations and by which mechanism intraorganellar domains are formed remains largely elusive. Here, we report on enzymes from the fungus Ustilago maydis , which accumulate inside of peroxisomal subdomains. We describe a short peptide motif (Thr-Ile-Ile-Val) sufficient to trigger focal localization. Mining for proteins with similar motifs uncovered several peroxisomal matrix proteins that accumulate in intraorganellar foci. These foci are enriched in the enzyme urate oxidase – a typical constituent of the paracrystalline core of peroxisomes. Upon peroxisome proliferation uneven distribution of focal structures results in the formation of peroxisome subpopulations with different protein content. The underlying principle of subdomain formation is evolutionary conserved in mammalian peroxisomes and formation of similar foci was also observed inside of mitochondria. We propose that peroxisomal proteins show an individual propensity to self-assemble. This formation of protein aggregates appears to be a ubiquitous driving force to spatially organize the peroxisomal proteome.

The phytopathogenic smut fungus Ustilago maydis is a versatile model organism to study plant pathology, fungal genetics, and molecular cell biology. Here, we report several strategies to manipulate the genome of U. maydis by the CRISPR/Cas9 technology. These include targeted gene deletion via homologous recombination of short double-stranded oligonucleotides, introduction of point mutations, heterologous complementation at the genomic locus, and endogenous N-terminal tagging with the fluorescent protein mCherry. All applications are independent of a permanent selectable marker and only require transient expression of the endonuclease Cas9hf and sgRNA. The techniques presented here are likely to accelerate research in the U. maydis community but can also act as a template for genome editing in other important fungi.

Phosphorylation is a key event in many cellular processes like cell cycle, transformation of environmental signals to transcriptional activation or polar growth. The chemical genetics approach can be used to analyse the effect of highly specific inhibition in vivo and is a promising method to screen for kinase targets. We have used this approach to study the role of the germinal centre kinase Don3 during the cell division in the phytopathogenic fungus Ustilago maydis. Due to the easy determination of the don3 phenotype we have chosen this approach for a genetic course for M.Sc. students and for IMPRS (International Max-Planck research school) students. According to the principle of “problem-based learning” the aim of this two-week course is to transfer knowledge about the broad spectrum of kinases to the students and that the students acquire the ability to design their own analog-sensitive kinase of interest. In addition to these training goals, we benefit from these annual courses the synthesis of basic constructs for genetic modification of several kinases in our model system U. maydis.