Research Focus

Molecular assessment of microbial communities on herbal tissues

We are all aware that ecosystems consist of a variety of organisms that co-exist and also interact with each other on multiple levels. Some of these are obvious interactions as herbivory or predation, clearly visible to everyone. But looking at the hidden microbiotic scale, these interactions are almost countless, nonetheless essential for all living organisms and currently not well understood. Most of the present knowledge is based on studies that focus on organisms that are cultivateable. Further, most studies are restricted on specific taxonomic groups.

In several projects, we aim to avoid these cultivation and taxonomic restrictions by examination of microbiotic diversity associated with plants with novel molecular sequencing techniques. With next-generation sequencers available, we are able to estimate microbiotic diversity and make suggestions about organismal abundances. In a preliminary, cultivation based study we were able to show that bacterial communities differ largely between flowers and leaves. We will expand this result by a variety of investigated plant species and looking at the microbiome dynamics from closed buds over flowers to seeds. Further, we will not only regard bacterial organisms, but also include fungi, plant and animal species in our investigation.

Referency library refinement and threshold-free DNA barcoding

Molecular approaches have gained a lot of importance in most biological fields, without which today’s research would be unimaginable. This also applies to genetic sequences used to identify and classify organisms, often refered to as DNA barcodes or phylogenetic markers. There are many restrictions, limiting the complete list of genes to a few markers appropriate to the task. However even for these markers, experts are discordant. There are taxonomic preferences, so that different phyla are investigated with different markers. These preferences are partially due to technical advantages and partially due to historic reasons. In every case, a reference sequence libary is needed against which the sequences of interest are queried. The current trend is that genetic data is becoming very abundant in public databases, available for this purpose. While this is a huge step into the right direction, this has some serious drawbacks: due to automation procedures and large scale approaches, a quite large amount of the data is inaccurate. Resulting problems for species identification occur when species or taxonomic associations have been wrongly assigned and/or sequences are of low quality due to sequencing errors. It is obvious that erroneos references inevitably lead to wrong conclusions in the following research projects.

We are currently working against this problem. By the combination of several clustering algorithms with filtering steps and outlier detection methods, we develop a bioinformatical pipeline to exclude low quality data, and to reclassify wrongly assigned sequences to their true taxonomic position. The idea behind this is to step away from narrow species definitions, but to develop a pragmatic reference library with molecular operational taxonomic units. We further focus on making this procedure adaptable to a variety of markers, ranging from very conserved (e.g. 16S) to very variable regions (e.g. ITS2). Due to that we are confident to achieve applicability for all major phyla.

Ongoing collaborations

  • Functional elucidation of WISP1 protein function within the musculoskeletal system (with K. Schlegelmilch and N. Schütze)
  • Impact of SNPs on collagen structural stability (with S. Rost et al.)
  • Functional meta-analyses of amphibian communities (with R. Ernst)
  • Poison dart frog female mating behavior (with I. Meuche)
  • Tardigrade culturing (with Dept. of Bioinformatics)