Dominik Frank’s research while affiliated with University of Tübingen and other places

Geometric morphometry has evolved as a powerful tool to unravel species delimitations within the genus Aristolochia. A survey conducted on the variation of Aristolochia trilabiata flowers and leaves revealed an overlooked entity in its affinity, which is newly described herein as Aristolochia franzii. The new species differs from its relative A. trilabiata by various floral characters, notably the presence of papillae on both the upper and lower limb zones, the presence of a well defined medial upper limb zone, the number of veins on the lateral upper limb zone, a considerably shorter tube, and the relative position of upper and lower limb zones. Furthermore, the leaf shape of A. franzii is cordiform-elongate to hastate, compared to the consistently more compact and shorter cordiform leaves of A. trilabiata. So far, A. franzii has been recorded from Northern Brazil (Amazonas) and French Guiana. An illustration of the diagnostic characters and comparison with A. trilabiata is provided, as well as the geographic distribution and a preliminary assessment of the conservation status.

Aristolochia holostylis has long been construed as constituting a single widespread and variable species, ranging across Paraguay, Bolivia, and Brazil. However, our morphological analysis shows that populations in Bolivia, Paraguay and western Brazil differ by several constant characters of the perianth and indumentum from those in northeastern Brazil. The type specimen possesses the typical morphology of the latter populations; thus, they retain the name A. holostylis, whereas the former populations are recognized here as belonging to a new species, A. pyriflora. We provide new descriptions for both species, along with morphological comparison between them, illustration of diagnostic characters and geographic distributions, and assessment of their respective conservation statuses. The often overlooked infraspecific name Holostylis reniformis f. minor is placed in the synonymy of A. pyriflora.

Septins are part of the cytoskeleton and polymerize into non-polar filaments of heteromeric hexamers or octamers. They belong to the class of P-loop GTPases but the roles of GTP binding and hydrolysis on filament formation and dynamics are not well understood. The basic human septin building block is the septin rod, a hetero-octamer composed of SEPT2, SEPT6, SEPT7, and SEPT9 with a stoichiometry of 2:2:2:2 (2-6-7-9-9-7-6-2). Septin rods polymerize by end-to-end and lateral joining into linear filaments and higher ordered structures such as rings, sheets, and gauzes. We purified a recombinant human septin octamer from E. coli for in vitro experimentation that is able to polymerize into filaments. We could show that the C-terminal region of the central SEPT9 subunit contributes to filament formation and that the human septin rod decreases the rate of in vitro actin polymerization. We provide further first kinetic data on the nucleotide uptake- and exchange properties of human hexameric and octameric septin rods. We could show that nucleotide uptake prior to hydrolysis is a dynamic process and that a bound nucleotide is exchangeable. However, the hydrolyzed γ-phosphate is not released from the native protein complex. We consequently propose that GTP hydrolysis in human septins does not follow the typical mechanism known from other small GTPases.

Septins are part of the cytoskeleton and polymerize into non-polar filaments of heteromeric hexamers or octamers. They belong to the class of P-loop GTPases but the roles of GTP binding and hydrolysis on filament formation and dynamics are not well understood. The basic human septin building block is the septin rod, a hetero-octamer composed of SEPT2, SEPT6, SEPT7, and SEPT9 with a stoichiometry of 2:2:2:2 (2-7-6-9-9-6-7-2). Septin rods polymerize by end-to-end and lateral joining into linear filaments and higher ordered structures such as rings, sheets, and gauzes. We purified a recombinant human septin octamer from E. coli for in vitro experimentation that is able to polymerize into filaments. We could show that the C-terminal region of the central SEPT9 subunit contributes to filament formation and that the human septin rod decreases the rate of in vitro actin polymerization. We provide further first kinetic data on the nucleotide uptake- and exchange properties of human hexameric and octameric septin rods. We could show that nucleotide uptake prior to hydrolysis is a dynamic process and that a bound nucleotide is exchangeable. However, the hydrolyzed γ-phosphate is not released from the native protein complex. We consequently propose that GTP hydrolysis in human septins does not follow the typical mechanism known from other small GTPases.