Fakultät für Chemie und Pharmazie - Digitale Hochschulschriften der LMU - Teil 04/06 Archiv

To become biologically active, a protein must fold into a distinct three-dimensional structure. Many non-native proteins require molecular chaperones to support folding and assembly. These molecular chaperones are important for de novo protein folding as well as refolding of denatured proteins under stress conditions. A certain subset of chaperones, the chaperonins, are required for the folding of the enzyme ribulose-1,5-bisphosphate carboxylase/oxygenase (Rubisco); furthermore, correct folding of Rubisco is also aided by the Hsp70 chaperone system. Rubisco catalyzes the initial step of CO2 assimilation in the Calvin-Benson-Bassham (CBB) cycle. Unfortunately, this enzyme is extremely inefficient, not only does it exhibit a slow catalytic rate (three CO2 molecules fixed per second per Rubisco) but it also discriminates poorly between the assimilation of CO2 and O2 to its sugar-phosphate substrate ribulose-1,5-bisphosphate (RuBP), the latter resulting in loss of photosynthetic efficiency. Due to these inefficiencies, carbon fixation by Rubisco is the rate limiting step of the CBB cycle. Photosynthetic organisms must produce tremendous amounts of Rubisco to alleviate these shortcomings; therefore significant quantities of nitrogen stores are invested in the production of Rubisco making Rubisco the most abundant protein on earth. These drawbacks of Rubisco have important implications in increasing CO2 concentrations and temperatures in the context of global warming. The ability to engineer a more efficient Rubisco could potentially reduce photosynthetic water usage, increase plant growth yield, and reduce nitrogen usage is plants. However, eukaryotic Rubisco cannot fold and assemble outside of the chloroplast, hindering advancements in creating a more efficient Rubisco. Form I Rubisco, found in higher plants, algae, and cyanobacteria, is a hexadecameric complex consisting of a core of eight ~50 kDa large subunits (RbcL), which is capped by four ~15 kDa small subunits (RbcS) on each end. The discovery of a Rubisco-specific assembly chaperone, RbcX, has lead to a better understanding of the components necessary for the form I Rubisco assembly process. RbcX is a homodimer of ~15 kDa subunits consisting of four α- helices aligned in an anti-parallel fashion along the α4 helix. RbcX2 functions as a stabilizer of folded RbcL by recognizing a highly conserved C-terminal sequence of RbcL: EIKFEFD, termed the C-terminal recognition motif. As has been demonstrated by studies of cyanobacterial Rubisco, de novo synthesized RbcL is folded by the chaperonins, whereupon RbcX2 stabilizes the folded RbcL monomer upon release from the folding cavity and then assists in the formation of the RbcL8 core. RbcX2 forms a dynamic complex with RbcL8 and as a result, RbcX2 is readily displaced by RbcS docking in an ATP-independent manner, thereby creating the functional holoenzyme. However, the exact mechanism by which RbcS binding displaces RbcX2 from the RbcL8 core is still unknown. Furthermore, though much advancement has been made in the understanding of form I Rubisco folding and assembly, an exact and detailed mechanism of form I Rubisco assembly is still lacking. The highly dynamic complex of RbcL/RbcX is critical for the formation of the holoenzyme; however it has hindered attempts to characterize critical regions of RbcL that interact with the peripheral regions of RbcX2. An important observation arose when heterologous RbcL and RbcX2 components interacted; a stable complex could form enabling in depth characterization of the RbcL/RbcX2 interaction. In the present study, the detailed structural mechanism of RbcX2-mediated cyanobacterial form I Rubisco assembly is elucidated. To obtain molecular insight into the RbcX2-mediated assembly process of cyanobacterial form I Rubisco, cryo-EM and crystallographic studies in concert with mutational analysis were employed by taking advantage of the high affinity interaction between RbcL and RbcX2 in the heterologous system (Synechococcus sp. PCC6301 RbcL and Anabaena sp. CA RbcX2). Structure guided mutational analysis based on the 3.2 Å crystal structure of the RbcL8/(RbcX2)8 assembly intermediate were utilized to determine the precise interaction site between the body of RbcL and the peripheral region of RbcX2. From these studies a critical salt bridge could be identified that functions as a guidepoint for correct dimer formation, and it was observed that RbcX2 exclusively mediates Rubisco dimer assembly. Furthermore, the mechanism of RbcX2 displacement from the RbcL8 core by RbcS binding was elucidated as well as indications of how RbcS docking on the RbcL8 core is imperative for full form I Rubisco catalytic function by stabilizing the enzymatically competent conformation of an N-terminal loop of Rubisco termed the ‘60ies loop’. Finally, initial attempts in in vitro reconstitution of eukaryotic Rubisco are reported along with the characterization of Arabidopsis thaliana RbcX2 binding to the C-terminal recognition motif of the Rubisco large subunit from various species.

In order to deliver drugs to diseased cells nanoparticles featuring controlled drug release are developed. Controlled release is of particular importance for the delivery of toxic anti-cancer drugs that should not get in contact with healthy tissue. To evaluate the effectivity and controlled drug release ability of nanoparticles in the target cell, live-cell imaging by highly-sensitive fluorescence microscopy is a powerful method. It allows direct real-time observation of nanoparticle uptake into the target cell, intracellular trafficking and drug release. With this knowledge, existing nanoparticles can be evaluated, improved and more effective nanoparticles can be designed. The goal of this work was to study the internalization efficiency, successful drug loading, pore sealing and controlled drug release from colloidal mesoporous silica (CMS) nanoparticles. The entire work was performed in close collaboration with the group of Prof. Thomas Bein (LMU Munich), where the nanoparticles were synthesized. To deliver drugs into a cell, the extracellular membrane has to be crossed. Therefore, in the first part of this work, the internalization efficiency of PEG-shielded CMS nanoparticles into living HeLa cells was examined by a quenching assay. The internalization time scales varied considerably from cell to cell. However, about 67% of PEG-shielded CMS nanoparticles were internalized by the cells within one hour. The time scale is found to be in the range of other nanoparticles (polyplexes, magnetic lipoplexes) that exhibit non-specific uptake. Besides internalization efficiency, successful drug loading and pore sealing are important parameters for drug delivery. To study this, CMS nanoparticles were loaded with the anti-cancer drug colchicine and sealed by a supported lipid bilayer using a solvent exchange method (additional collaboration with the group of Prof. Joachim Rädler, LMU). Spinning disk confocal live-cell imaging revealed that the nanoparticles were taken up into HuH7 cells by endocytosis. As colchicine is known to exhibit toxicity towards microtubules, the microtubule network of the cells was destroyed within 2 h of incubation with the colchicine-loaded lipid bilayer-coated CMS nanoparticles. Although successful drug delivery was shown, it is necessary to develop controlled local release strategies. To achieve controlled drug release, CMS nanoparticles for redox-driven disulfide cleavage were synthesized. The particles contain the ATTO633-labeled amino acid cysteine bound via a disulfide linker to the inner volume. For reduction of the disulfide bond and release of cysteine, the CMS nanoparticles need to get into contact with the cytoplasmic reducing milieu of the target cell. We showed that nanoparticles were taken up by HuH7 cells via endocytosis, but endosomal escape seems to be a bottleneck for this approach. Incubation of the cells with a photosensitizer (TPPS2a) and photoactivation led to endosomal escape and successful release of the drug. In addition, we showed that linkage of ATTO633 at high concentration in the pores of silica nanoparticles results in quenching of the ATTO633 fluorescence. Release of dye from the pores promotes a strong dequenching effect providing an intense fluorescence signal with excellent signal-to-noise ratio for single-particle imaging. With this approach, we were able to control the time of photoactivation and thus the time of endosomal rupture. However, the photosensitizer showed a high toxicity to the cell, due to its presence in the entire cellular membrane. To reduce cell toxicity induced by the photosensitizer and to achieve spatial control on the endosomal escape, the photosensitizer protoporphyrin IX (PpIX) was covalently surface-linked to the CMS nanoparticles and used as an on-board photosensitizer (additional collaboration with the groups of Prof. Joachim Rädler and Prof. Heinrich Leonhardt, both LMU). The nanoparticles were loaded with model drugs and equipped with a supported lipid bilayer as a removable encapsulation. Upon photoactivation, successful drug delivery was observed. The mode of action is proposed as a two step cascade, where the supported lipid bilayer is disintegrated by singlet oxygen in a first step and the endosomal membrane ruptures enabling drug release in a second step. With this system, stimuli-responsive and controlled, localized endosomal escape and drug release is achieved. Taken together, the data presented in this thesis show that real-time fluorescence imaging of CMS nanoparticles on a single-cell level is a powerful method to investigate in great detail the processes associated with drug delivery. Barriers in the internalization and drug delivery are detected and can be bypassed via new nanoparticle designs. These insights are of great importance for improvements in the design of existing and the synthesis of new drug delivery systems.

1. It was demonstrated that the two-fold deprotonation of benzofuran-3-yl- and benzothiophen-3-ylanilines using nBuLi, followed by transmetalation using CuCl·2LiCl afforded cyclic amidocuprates which were readily oxidized using 2,3,5,6-tetrachlorobenzo-p-quinone (“chloranil”) to furnish new tetracyclic benzofuro- and benzothieno[2,3-b]indoles. 2. A method was developed, starting from heteroaryl organometallics, for the synthesis of non-conjugated heteroaromatic benzylic zinc compounds which could be reacted with a wide range of electrophiles. 3. The preparation of mixed aminals from Mannich’s cation and a range of metal amides was performed and their subsequent ionization to (methylene)iminium ions using trifluoroacetic anhydride was achieved. Trapping of these iminium ions with a range of organometallic substrates in a one-pot fashion led to the efficient formation of sterically hindered tertiary benzylic and phenethylic amines.

Topic I Molecular basis of RNA polymerase III transcription repression by Maf1 RNA polymerase III (RNAP III) is a conserved 17-subunit enzyme that transcribes genes encoding short untranslated RNAs such as transfer RNAs (tRNAs) and 5S ribosomal RNA (rRNA). These genes are essential and involved in fundamental processes like protein biogenesis; hence RNAP III activity needs to be tightly regulated. RNAP III is repressed upon stress and this is regulated by Maf1, a protein conserved from yeast to humans. Many stress pathways were shown to converge on Maf1 and result in its phosphorylation, followed by its nuclear import and eventual repression of RNAP III activity. However, the molecular mechanisms of this repression activity were not known at the beginning of these studies. This work establishes the mechanism of RNAP III specific transcription repression by Maf1. The crystal structure of Maf1 was solved. It has a globular fold with surface accessible NLS sequences, which sheds new light on already published results and explains how stress-induced phopshorylation leads to import of Maf1 into the nucleus. Additionally, cryo EM studies and competition assays show that Maf1 binds RNAP III at its clamp domain and thereby induces structural rearrangements of RNAP III, which inhibits the interaction with Brf1, a subunit of the transcription initiation factor TFIIIB. This specifically impairs recruitment of RNAP III to its promoters and implies that Maf1 is a repressor of transcription initiation. Competition and transcription assays show that Maf1 also binds RNAP III that is engaged in transcription, leaving RNAP III activity intact but preventing re-initiation. Topic II Structure of human mitochondrial RNA polymerase The nuclear-encoded human mitochondrial RNAP (mitoRNAP) transcribes the mitochondrial genome, which encodes rRNA, tRNAs and mRNAs. MitoRNAP is a single subunit (ss) polymerase, related to T7 bacteriophage and chloroplast polymerases. All share a conserved C-terminal core, whereas the N-terminal parts of mitoRNAP do not show any homology to other ss RNAPs. Unlike phage RNAPs, which are self-sufficient, human mitoRNAP needs two essential transcription factors for initiation, TFAM and TFB2M. Both of these factors are likely to control the major steps of transcription initiation, promoter binding and melting. Thus human mitoRNAP has evolved a different mechanism for transcription initiation and exhibits a unique transcription system. Structural studies thus far concentrated on the nuclear enzymes or phage RNAPs, whereas the structure of mitochondrial RNA polymerase remained unknown. The structural organization of human mitoRNAP and the molecular mechanisms of promoter recognition, binding and melting were subject of interest in these studies. In this work the crystal structure of human mitoRNAP was solved at 2.4 Å resolution and reveals a T7-like C-terminal catalytic domain, a N-terminal domain that remotely resembles the T7 promoter-binding domain (PBD), a novel pentatricopeptide repeat (PPR) domain, and a flexible N-terminal extension. MitoRNAP specific adaptions in the N-terminus include the sequestering of one of the key promoter binding elements in T7 RNAP, the AT-rich recognition loop, by the PPR domain. This sequestration and repositioning of the N-terminal domain explain the need for the additional initiation factor TFAM. The highly conserved active site within the C-terminal core was observed to bind a sulphate ion, a well known phosphate mimic, and thereby suggests conserved substrate binding and selection mechanisms between ss RNAPs. However, conformational changes of the active site were observed due to a movement of the adjacent fingers subdomain. The structure reveals a clenching of the active site by a repositioned fingers subdomain and an alternative position of the intercalating -hairpin. This explains why the conserved transcription factor TFB2M is required for promoter melting and initiation. A model of the mitochondrial initiation complex was build to further explore the initiation mechanism, and to rationalize the available biochemical and genetic data. The structure of mitoRNAP shows how this enzyme uses mechanisms for transcription initiation that differ from those used by phage and cellular RNAPs, and which may have enabled regulation of mitochondrial gene transcription and adaptation of mitochondrial function to changes in the environment.