Browsing by Person "Kuhn, Andreas"

Now showing 1 - 16 of 16

Biogenese und Virusassembly des filamentösen Coliphagen M13
(2012) Ploß, Martin; Kuhn, Andreas
Taxonomically, the bacteriophage M13 is assigned to the single-stranded DNA phage and belongs to the family of Inoviridae. For propagation the Gram-negative bacteria Escherichia coli with F-pili is required. The host cell is not lysed by the phage. New findings about the M13 phage biogenesis are presented here within four essential areas of the M13 phage cycle concerning the sections infection, assembly, and phage secretion. Phage adsorption experiments in which the host bacterium E. coli K38 was infected by M13 phage showed that the phage adsorption to the cells takes place within the first 5 minutes and because of a limitation of F-pili per cell a maximum of 7 phages per cell were found to be adsorbed. The insertion of the phage coat protein gp9 into the cytoplasmic membrane of the host cell was verified by the periplasmic location of antigenic epitopes introduced into the N-terminal domain of gp9. The membrane insertion of gp9 was found to depend on the host protein YidC. Plasmid-encoded gp9 exhibiting antigenic epitopes at the N-terminal domain did not interfere with the assembly of new progeny phage. Therefore, the development of a phage display system with gp9 by introducing short peptide sequences (17 ? 36 amino acids) is feasible. After overexpression of gp1/11 assembly complexes in E. coli and size exclusion chromatography, respectively, the complex was characterized and a molecular weight of ~ 300 kDa was assigned. Examinations of the purified gp1/11 assembly complexes by transmission electron microscopy (TEM) revealed ring-like structures with ~ 7 ? 8 nm in inner diameter and ~ 11 ? 12 nm in outer diameter. The investigation of M13 wild-type infection showed that the secretion of new progeny phage starts after a short lag period (eclipse). An infected E. coli cell secreted upto 925 progeny in a time period of 115 minutes which corresponds to an average of 7 secreted phages per minute. The generation time of the infected E. coli K38 cells rose from 24 minutes to 48 minutes. Experiments were carried out with genetically manipulated phages which were hindered to synthesize the major coat protein gp8 in the host cell by a nonsense mutation in the phage genome. Therefore, phage replication was only observed in host cells bearing plasmid encoding gene 8. Since the quantity of the protein was limited the lag period (eclipse) was extended to 12 minutes and the efficiency of phage secretion was decreased to about 2 phages per minute. The M13 phage secretion from infected E. coli cells was visualized by atomic force microscopy (AFM). The identity of the phage was verified by labeling with protein-A conjugated gold and transmission electron microscopy (TEM). The secretion of M13 progeny was first observed at the cell poles of E. coli and then spreaded within 4 minutes along the cell surface. After 16 minutes the secretion was observed over the entire cell surface.
Functional and structural studies of a C-terminally extended YidC
(2015) Seitl, Ines; Kuhn, Andreas
Members of the YidC/Oxa1/Alb3 protein family catalyze the insertion of integral membrane proteins into the lipid bilayer of the bacterial plasma membrane (YidC), the inner mitochondrial membrane (Oxa1), and the chloroplast thylakoid membrane (Alb3) (Saller et al., 2012; Dalbey et al., 2014). The insertase homologs are comprised of a conserved core region of 5 transmembrane domains, but are provided with additionally flanking N- and C-terminal regions of variable lengths and functions. The Gram-negative YidC is characterized by an additional N-terminal domain, while Gram-positive bacteria, mitochondria and plastids developed C-terminally extended insertase-domains. These domains are involved e.g. in direct interaction with ribosomes and facilitate a functional overlap with the co-translational SRP-targeting pathway. An extended C-terminal highly positively charged tail region was also found in the YidC homologs of the Gram-negative marine bacteria Rhodopirellula baltica and Oceanicaulis alexandrii, but not in Escherichia coli. The primary subject of this work was to characterize and analyze in detail the C-terminally extended YidC chimera, composed of the E. coli YidC and the C-terminally extended domains of the marine YidC homologs. Biochemical binding assays with the purified YidC proteins and isolated, vacant E. coli 70S ribosomes showed that the C-tails mediate specific binding to ribosomes independently of the translational state of the ribosome. Furthermore, a ribosome-bound insertase complex was visualized by cryo-electron microscopy. The enhanced affinity of the C-terminally extended YidC was used to isolate stable complexes with stalled ribosomes, carrying a nascent polypeptide chain of a YidC substrate protein (MscL). The cryo-EM structure of a YidC-ribosome nascent chain complex (RNC) was solved to a 8,6 Å resolution and allowed the visualization of the nascent chain from the peptidyl transferase center through the ribosomal exit tunnel into the YidC density. The structure revealed the helix H59 of the 23S rRNA and the two ribosomal proteins L24 and L29 as the major contacts sites of YidC at the ribosomal tunnel exit. Pull down assays confirmed a significantly interaction of the C-terminal ribosome binding domain and the ribosomal protein L29, while L24 seems to be a universal contact site for the YidC-insertase core domain. Strikingly, the cryo-EM structure clearly showed a single monomer of YidC bound to the translating ribosome. This suggests that monomeric YidC might be the minimal functional unit for YidC-dependent, co-translational insertion of inner membrane proteins. In addition to the in vitro tests, a possible role of the C-terminal YidC extensions in co-translational protein targeting was tested in vivo in E. coli. For that purpose the targeting and localization of the SRP-dependent YidC-substrate protein MscL (Facey et al., 2007) was investigated as a GFP fusion protein via fluorescence microscopy. In addition, the proper membrane insertion of MscL was analyzed in radioactive pulse chase experiments via AMS gel shift assays, either in the absence of a functional SRP or SRP receptor (FtsY). Both in vivo assays clearly showed that the C-terminal ribosome binding domain of the R. baltica YidC homolog can partially substitute for the SRP receptor function in E. coli, while the cytosolic signal recognition particle is still required for correct insertion of the MscL protein. Therefore, a new co-translational targeting and insertion model of YidC-only substrates was proposed. This works also highlights evolutionary aspects of the accessory YidC domains and indicates that the C-terminal extended tail of YidC in the planctomycete group may be an ancestral remnant of a primordial translocation system operating without a typical SRP receptor. The second part focuses on the interaction of the signal recognition particle with SRP signal sequences. Isolated mutant signal sequence peptides were used to determine the specificity of SRP recognition in proteins. The interaction studies were established in an in vitro system and binding affinities of purified SRP to the isolated signal sequence peptides were determined via microscale thermophoresis (MST). A short sequence of 27 amino acid residues at the very N-terminal tail of the large cytoplasmic domain of KdpD was identified as a SRP signal sequence. Furthermore, a direct influence of the amino acid composition in the signal peptide on its SRP binding affinity in vitro was demonstrated. This confirms a low influence of an altered charge in the N-terminal region while mutations in the hydrophobic core region causes significantly reduced binding affinities to SRP. Taken together, this study contributes to the understanding of the molecular mechanisms of co-translational membrane protein biogenesis in bacteria.
Funktion und Dynamik eines gemeinsamen Insertionskomplexes der Sec-Translokase und YidC-Insertase in der bakteriellen Membran
(2020) Steudle, Anja; Kuhn, Andreas
YidC/Oxa1/Alb3-insertases and the Sec-translocase are conserved across all three kingdoms of life and constitute the most important pathway for integral proteins into cell membranes and membranes of eukaryotic organelles. The insertion of membrane proteins into the inner membrane of Gram-negative bacteria occurs mainly via the SecYEG-translocase and the YidC-insertase acting independently or in cooperation. For the cooperative insertion a close contact between SecY and YidC is assumed. Previous interaction-studies and a recently solved low-resolution structure of the so-called holo-translocon (14 Å) indicate a contact between the lateral gate of SecY and the hydrophobic substrate slide of YidC. Which specific domains of YidC and SecY thereby interact directly with each other was unknown so far. The aim of this study was to describe the contact between SecY and YidC in more detail. A high affinity for the interaction of the two proteins in detergent and in DOPC-proteoliposomes was determined via FRET measurements with fluorescently labeled SecY and YidC. For the stoichiometric ratio of the SecY/YidC-interaction a factor of one was calculated. To identify the specific contacts between SecY and YidC in vivo disulphide cross-linking experiments were performed. Direct interactions between the transmembrane domain (TM) 3 and TM8 of the SecY lateral gate and TM3 and TM5 of the hydrophobic slide of YidC were found, respectively. Furthermore, a YidC mutant with five serine substitutions, which was unable to rescue a YidC depletion strain, was investigated. Even though the serine positions are located in the middle and the periplasmic half of the hydrophobic slide of YidC and four of the positions are identical with substrate contact sites, no inhibition of insertion for the YidC-dependent substrates M13 procoat and Pf3 coat by the 5S mutant compared to the wildtype YidC was observed. For the YidC-only pathway a minimum of hydrophobicity seems to be required sufficient to allow the insertion of these substrates. In vitro FRET measurements showed an impaired interaction between SecY and the YidC 5S mutant and confirmed once again an involvement of the hydrophobic slide in the SecY/YidC-contact. Based on the cross-linking contacts and the results of the FRET measurements a possible model of the SecY/YidC-contact was established, which shows the SecY lateral gate vis-à-vis of the hydrophobic slide and the hydrophilic groove between TM3 and TM5 of YidC generating a combined SecY/YidC-cavity. Taken together, the present study provides further evidence that the lateral gate of the Sec-translocase directly interacts with the hydrophobic slide of YidC. In a further project, a SecY-YidC fusion protein was cloned to ensure the two proteins are in close proximity, the correct orientation and proper stoichiometry after reconstitution into proteoliposomes. For a collaboration with the ETH Zürich, proteoliposomes hosting the fusion protein, SecYEG, YidC or SecYEG and YidC together were prepared by myself in Hohenheim. The stepwise insertion of the Sec/YidC-dependent substrate LacY into these proteoliposomes was observed by a collaborating group of the ETH Zürich using AFM–based single-molecule force spectroscopy. The insertion of LacY was observed for the different cases but for the fusion protein and SecYEG combined with YidC the insertion process is dominated by the Sec-translocase, whereas YidC probably only has a supporting function in the folding of the protein.
Funktionelle Untersuchung der Sensorkinase KdpD von Escherichia coli mit Hilfe verschiedener KdpD-Deletionsmutanten
(2007) Rothenbücher, Marina; Kuhn, Andreas
The high affinity K+ transport system KdpFABC is one of several uptake systems that accumulate K+ in Escherichia coli. Expression of the kdpFABC operon is under control of the regulatory proteins KdpD and KdpE, which constitute a typical sensor kinase/response regulator system. KdpD is an integral protein of the cytoplasmic membrane. The N-terminal domain, the 4 helices and 200 amino acids of the cytoplasmic C-terminal domain are accredited to be involved in the signal input function. Surprisingly, a mutant (KdpD-C) lacking the N-terminal domain, helix 1 and 2 is a functional K+ sensor, which is able to detect the changes in K+ concentration in the medium. To investigate which parts of the KdpD protein are essential for signal transduction, various truncated KdpD variants were constructed and analyzed. The results show that the fragment C499-894, which contains only the cytoplasmic C-terminal domain of KdpD, is able to recognise the increase in K+ concentration in medium and reduce the level of activity. This mini sensor is also able to discriminate between Li+, Rb+ and K+ ions like the wild-type KdpD. A plasmid coding for this mini sensor allows a kdpD deletion strain to grow under K+-limited conditions in medium. Presumably, the signal perceived by KdpD is in the periplasm, but how is the signal transmitted to the cytoplasmic domain of KdpD that controls activity? Perhaps KdpD is not the direct sensor, an additional component in the membrane might sense the signal and communicate with KdpD. Further research with sodium carbonate extraction to determine the membrane localisation of C499-894 showed the protein mainly found in the supernatant, suggesting C499-894 is a soluble protein, although C499-894 could be attached to the membrane to come in contact with an unknown membrane protein. Looking for the unknown component two protein candidates were found by co-purification, the cytochrome oxidase A (CyoA) and the glucosamine-6-phosphate synthase (GlmS). To find an interaction between these proteins and KdpD more research has to be done.
In vitro Interaktionsstudie der bakteriellen Insertase YidC von Escherichia coli mit seinem Substrat Pf3 coat Protein durch Fluoreszenzspektroskopie
(2011) Winterfeld, Sophie; Kuhn, Andreas
With fluorescence spectroscopy the insertion of Pf3 coat protein into YidC proteoliposomes could be characterized. Therefore the binding could be determined by fluorescence titration and with help of single-molecule spectroscopy the insertion process of Pf3 coat protein was analyzed with FRET. Thus the distances of the proteins were determined and the time dependent insertion was observed.
Die Insertion des „minor coat“ Proteins G3P des Bakteriophagen M13 in die innere E. coli Membran benötigt die Insertase YidC und die Translokase SecYEG.
(2021) Kleinbeck, Farina; Kuhn, Andreas
The membrane of every cell forms a spacial limitation for this smallest unit of a life form. Such a very simple unicellular life form is also the Gram-negative bacterium Escherichia coli (E. coli) and is therefore a valid model organism for a living cell. Due to the inner membrane the cellular components are held together in close proximity and are separated from the extracellular environment. Most substrates cannot pass the lipid bilayer, which forms the membrane, so an import and export system had to be developed to accomplish this. For these import and export systems, very complex, polytopic transmembrane protein complexes are needed. Examples are ion channels, ion pumps or large complexes through which energy production, secretion of toxins and the transfer of nutrients are catalysed. Moreover, proteins with functions in the periplasm or outer membrane must also travel from their site of synthesis in the cytoplasm to their destination. For these different processes proteins must be inserted into or translocated across the inner membrane. Of the total proteome in prokaryotes approximately 25 to 30% is either inserted into or secreted across the inner membrane. This work identified several components required for the insertion of the "minor coat" protein G3P of M13 bacteriophage. This protein is important for the assembly of the phage particle that occurs in the inner membrane. The outermost C-terminus of G3P is anchored in the inner membrane via a single transmembrane domain, while the bulk of the approximately 42 kDa protein is located in the periplasm. Using an N-terminal cleavable signal peptide, the major portion of G3P is translocated into the periplasm via SecYEG with the help of SecA and the membrane potential. Targeting, on the other hand, could not be clearly assigned to one of the known post- or co-translational pathways. Although contact via disulfide crosslink studies to Ffh, the protein component of the ribonucleoprotein SRP, was observed via stalled ribosome nascent chains (RNCs), insertion into the membrane in vivo was independent of Ffh. Even when the interaction between SecY and FtsY, the receptor for SRP at the membrane, was impaired, G3P was inserted via SecYEG. Although the chaperone SecB was able to bind to G3P in vitro, G3P inserted independently of SecB in vivo. For membrane incorporation of G3P, it was shown that YidC is required in vivo in addition to SecYEG. Disulphide crosslink studies demonstrated that G3P first contacts the plug domain TM2b and lateral gate (TM2a and TM7) via the signal peptide of G3P, and finally the C-terminal transmembrane domain of G3P contacts YidC via TM3 and TM5 of the hydrophilic slide. Based on these contact sites, a possible insertion model was confirmed, with SecY and YidC mediating defined steps in the insertion process, providing new insights into this largely unknown process.
Interaktion des Photosensors Ppr aus Rhodocista centenaria mit Proteinkomponenten der chemotaktischen Signaltransduktion
(2008) Kreutel, Sven; Kuhn, Andreas
Rhodocista centenaria (previously known as Rhodospirillum centenum) is a photosynthetic alpha-proteobacterium which exhibits a unique phototactic response in respect of the direction of light. In this work, the focus is on the potential photoreceptor Ppr and its C-terminal histidine kinase Pph to identify putative binding partners in the signal transduction pathway. The results of overexpression experiments with the Ppr-receptor or the Pph-domain in E. coli indicated that there may be an interaction between the photosensor and the chemotactic signalling pathway. Even cells expressing only small amounts of the R. centenaria proteins showed no chemotactic response at all, whereas uninduced cells exhibited normal chemotactic response on swarm plates as well as in capillary assays. To investigate whether the receptor interacts with components of the chemotactic pathway, the Ppr-protein and the Pph-domain as well as the chemotactic proteins CheW and CheAY of R. centenaria were heterologously expressed in E. coli and purified to homogeneity by affinity chromatography. Binding experiments were carried out by using an IAsys biosensor cuvette system. The results indicated that the kinase domain Pph binds to the chemotactic linker protein CheW with a dissociation constant of 0.13 ± 0.026 μM. Pull-down experiments were made to verify this finding and to investigate the role of ATP in the binding process. The results confirmed our previous observations but in contrast to the complex formation in the E. coli chemotactic pathway, the binding of the C-terminal histidine kinase to CheW was ATP-dependent. To study whether the kinase domain also binds CheW in vivo, expression and coelution experiments with tagged Pph-protein were carried out in R. centenaria. The findings suggest that the complex formation occurs in vivo as well as in vitro. From the data of the E. coli chemotactic complex formation it is well known that CheA is part of the trimeric complex which consists of MCP-receptors, CheW and CheA. To analyse this, pulldown experiments with all three proteins (Pph, CheW and CheAY) were performed. The results clearly showed a participation of CheAY in the formation of the complex with CheW and the kinase Pph. It is known that the photoreceptor Ppr is autophosphorylated during its light induced photocycle. We therefore examined whether the kinase domain is sufficient for this autophosphorylation reaction and whether CheAY could function as a phosphate acceptor. Our results confirmed the hypothesis that the kinase domain is sufficient for autophosphorylation and that the Pph-protein assists CheAY to take over phosphate groups. Taken together, the results in combination with data from the literature lead to a detailed working model for the function of the photoreceptor Ppr and the signal transduction pathway.
Interaktion des Portalring-Proteins gp20 des Bakteriophagen T4 mit Wirtsproteinen von Escherichia coli
(2012) Quinten, Tobias; Kuhn, Andreas
Bacteriophage T4 is composed of the three structural subunits i. e. capsid, tail and tail fibres. Because of its contractible tail T4 is a member of the Myoviridae. A characteristic feature of its morphogenesis is the membrane-associated assembly of the head structure during a wild-type infection of E. coli. The portal protein, gp20, and the phage-chaperone gp40 are used to form a membrane-bound complex. Most likely, also proteins from the phage-host E. coli are involved in this process. This complex is the starting point for the assembly of the head-related core and scaffold structure. The mature head detaches from the membrane and is filled with DNA thereafter. In the context of this doctoral thesis the role and function of the portal protein gp20 and its interactions with cellular proteins was analyzed. A His-tag was fused to gp20 and it was purified by nickel-affinity chromatography. Also, interactions with the cellular chaperones DnaK, GroEL, Tig and YidC and the phage-chaperone gp40 were detected after formaldehyde crosslinking. Further studies of the cellular localization showed that gp20 and the fusion protein gp20-GFP are membrane-bound. The importance of YidC and DnaK for this membrane-association was demonstrated by fluorescence microscopy. Phage propagation was not affected by YidC depletion, whereas the loss of DnaK led to a reduced propagation. Prohead-structures, that are an intermediate stage of the capsid assembly, were isolated in YidC-free E. coli-cells membrane-unbound when infected with a phage mutant. Previous studies had led to the isolation of different amber mutants. Mutant amber20E481 was used in this thesis to analyze the assembly process in more detail. Here, under non-permissive conditions a 14 amino acid shortened protein, gp20s, is synthesized. Despite the fact that the capsid assembly is blocked in the non-suppressor strain, the localization and expression of the truncated protein was comparable to the wild type gp20. Overexpressed and His-tagged gp20s was found to crosslink with YidC, GroEL and gp40. Structural studies with a transmission electron microscope showed, that mature proheads were found and these were not filled with phage DNA. Most probably, a malfunction of gp20s during DNA packaging accounts for this.
Membrane targeting and insertion of the sensor protein KdpD and the C-tail anchored protein SciP of Escherichia coli
(2019) Proß, Eva; Kuhn, Andreas
In E. coli, most inner membrane proteins are targeted in a co-translational manner by the universally conserved signal recognition particle (Bernstein et al. 1989; Valent et al. 1998; Schibich et al. 2016). SRP scans the translating ribosomes and binds with high affinity to an exposed SRP signal sequence, present in the nascent chain (Bornemann et al. 2008; Holtkamp et al. 2012; Saraogi et al. 2014). After targeting to the membrane-associated SRP receptor FtsY, the nascent membrane protein is forwarded to the Sec translocase or to the YidC insertase to be integrated into the bilayer (Miller et al. 1994; Cross et al. 2009; Welte et al. 2012; Akopian et al. 2013). In general, the targeting and insertion pathways of inner membrane proteins in E. coli are already well studied. However, there is a special class of proteins, the C-tail anchored proteins with only a few members in E. coli, whose insertion mechanisms are unknown in prokaryotes to date. To study those insertion mechanisms, the C-tail anchored protein SciP was used as a model protein. SciP from the enteroaggregative E. coli is a structural component of the type 6 secretion system and contains a transmembrane domain (TMD) at the extreme C-terminal part from amino acid 184 to 206. This results in a large N-terminal cytoplasmic domain of 183 amino acids. In E. coli, there is another protein, the potassium sensor protein KdpD which shares with SciP the commonality of a large N-terminal cytoplasmic domain. KdpD is a four-spanning membrane protein with the first TMD starting at amino acid position 400. For both proteins, with the TMD being located far away from the cytoplasmic N-terminal part, it was thought that they cannot use the co-translational SRP pathway. However, it was shown that KdpD is targeted co-translationally by SRP and a cytoplasmic targeting signal located between amino acids 22-48 was identified (Maier et al. 2008). In this study it was shown that the C-tail anchored protein SciP is also targeted early during translation by SRP. With fluorescence microscopy studies and sfGFP-SciP fusion constructs, two short hydrophobic regions in the N-terminal cytoplasmic domain (amino acids 12-20 and 62-71) were identified as being important for membrane targeting. With artificially stalled ribosomes exposing each of the targeting signal, microscale thermophoresis meausurements decoded that both signals bind to SRP and to a preincubated SRP-FtsY complex, mimicking the next targeting step. Cysteine-accessibilty assays demonstrated that SciP is the first described protein with two targeting signals since the deletion of one of the hydrophobic regions was compensated by the other remaining one in vivo. To decipher the crucial features of the novel cytoplasmic SRP signal sequences of KdpD and SciP alterations in the signal sequences were analyzed with fluorescence microscopy using sfGFP fusion constructs and microscale thermophoresis measurements using stalled ribosomes. These studies revealed that the novel signal sequences have to exceed a threshold level of hydrophobicity to be recognized and bound by SRP and target sfGFP to the membrane. In addition, three positively charged amino acids in the KdpD SRP signal sequence were identified to promote SRP binding. To characterize the binding mechanism of SRP to the signal sequences, in vitro disulphide cross-linking studies with synthesized KdpD22-48, SciP1-27 and SciP54-85 peptides were performed. All three peptides could be cross-linked to the hydrophobic groove of SRP formed by the M domain, which correlates with the binding of SRP to other substrates. Taken together, the results show that SRP binding is not limited to the TMDs of proteins. SRP is also able to recognize short hydrophobic stretches in the cytoplasmic domain of inner membrane proteins. Cysteine-accessibility assays with the C-tail anchored protein SciP decoded that not only SRP is involved in the delivery pathway but also the insertase YidC. With only 11 amino acids in the periplasmic domain SciP matches with the characteristics of other known YidC only substrates. By extending the C-tail of SciP it was found out that a critical length of 20 amino acids exists and that the exceed of this limit makes the insertion of SciP dependent on the Sec translocase. The studies with the extended C-tails of SciP helped to gain more general information about the YidC dependent insertion of proteins. The results obtained with the protein SciP are first indications about how the insertion of C-tail anchored proteins occurs in E. coli. It is assumed that the SRP system and the insertase YidC compensate the absence of the eukaryotic Get system, responsible for the insertion of eukaryotic tail-anchored proteins.
Membraneinbau von MscL und MscL-Mutanten aus Escherichia coli
(2012) Neugebauer, Stella; Kuhn, Andreas
About one third of all synthesized proteins in a cell are membrane proteins. To accomplish their function, it is important to ensure, that they safely reach their destination, insert efficiently into the membrane, where they fold into their correct tertiary structure. Previous studies have shown that various molecules are responsible for the targeting and insertion of membrane proteins in Escherichia coli that operate as individual modules. The mechanosensitive channel MscL is a pentameric complex in the cytoplasmic membrane of E. coli. By its action as a safety valve, MscL allows the adaption to hypoosmotic conditions of bacteria living under varying circumstances. The two transmembrane segments of the MscL monomer are connected by a periplasmic loop of 29 amino acid residues. In previous studies, the membrane insertion of MscL was analyzed in vivo in depletion strains and was monitored by modification of a single cysteine residue in the periplasmic domain of the MscL protein (Facey et al., 2007). The targeting of MscL to the inner membrane occurs in a cotranslational manner via the signal recognition particle (SRP). At the membrane, the MscL protein inserts independently of the membrane potential and the Sec-components SecAYEG, but requires YidC for the insertion process. The present thesis is about the molecular mechanisms regarding the decision whether the nascent polypeptide chain of MscL is recognized and bound by YidC or by the Sec-translocase. The periplasmic localized loop of MscL was altered by introducing negatively or positively charged residues as well as uncharged side chains and the effects on the translocation were investigated. Translocation of the periplasmic domain of MscL was detected using AMS-derivatization (4-acetamido-4´-maleimidylstilbene-2, 2´-disulfonic acid) of a single cysteine residue. The extension of the loop region by one, two or three negatively charged residues (aspartic acid residues) made the insertion of MscL dependent on the membrane potential and the Sec translocon. The requirement of SecYE was gradually affected by increasing the number of charged residues. Efficient translocation of the periplasmic loop with three additional uncharged (asparagines) residues also required the Sec-complex. The insertion of these MscL mutants was independent on the SecA component, but all the investigated mutants still showed a strict dependence on YidC. The ability of the altered MscL proteins to form functional pentameric channels was verified by growth tests and native gel electrophoresis. The presence of three additional positively charged arginine residues in the periplasmic domain inhibited MscL insertion into the lipid bilayer as well as the mutant with five additional negatively charged aspartic acid residues. As a logical consequence, the expression of these two MscL proteins could not protect the cells from osmolysis within growth tests. The direct involvement of the membrane insertase YidC with MscL and the MscL mutants was corroborated with in vivo crosslinking. YidC interacts with both transmembrane regions of MscL. Earlier studies have shown that YidC makes contact with the Pf3 coat protein in the center of the membrane. Here, the same interaction sites of YidC were identified contacting MscL during its insertion. Besides considering the significance of YidC for efficient membrane insertion, the present work has demonstrated that YidC is also essential for oligomerization of MscL into a functional channel.
Membraninsertion des Phagenproteins M13 procoat in Lipidvesikel mit rekonstituiertem Escherichia coli YidC
(2011) Stiegler, Natalie; Kuhn, Andreas
Translocation of proteins across or into the cytoplasmic membrane of Escherichia coli is accomplished by several mechanisms. The cellular secretion machinery, the translocase SecYEG, mediates the transport of unfolded proteins into the periplasm with the help of the ATPase SecA or passes the membrane proteins for bilayer integration to the insertase YidC. Membrane insertion is catalysed by YidC, whereby the native conformation of the proteins in the lipid bilayer is achieved. The translocation of a few membrane proteins occurs Sec-independently solely with the help of the insertase YidC. One of these Sec-independent proteins is the major capsid protein of the bacteriophage M13. This protein is inserted as preprotein, termed M13 procoat, with the orientation Nin-Cin into the inner membrane and a central loop domain located in the periplasm. This process is catalysed by the electrochemical membrane potential and YidC. M13 procoat is then processed by the leader peptidase to its mature form, M13 coat (orientation Nout-Cin). In the present thesis an analysis of the different transport systems of the inner membrane is performed using the example of the M13 procoat protein and its mutants. One mutant is the procoat H5EE which has 2 additional acidic residues introduced between residues +2 and +3. The insertion of this mutant requires the Sec translocase and strictly depends on the electrochemical potential. Membrane insertion of M13 procoat and derived proteins into the cytoplasmic membrane was followed in an in vitro reconstitution and translocation system. Therefore, all components of the Sec translocase (SecYEG and SecA), the insertase YidC and the different procoat proteins were purified and tested with the in vitro translocation system. Reconstitution of YidC into phospholipid vesicles depended on the lipid composition for its orientation. The cytoplasmic-out orientation corresponds to the active topology in E. coli where both termini are located in the cytoplasm. Certain lipid compositions caused the inversed orientation, which affected the catalytic activity of the reconstituted insertase. The procoat mutants H5 und H5EE were membrane inserted only in the presence of reconstituted YidC. Both proteins inserted efficiently into the vesicles with the periplasmic loop in the interior of the vesicles like the mutant PClep of procoat H5 with the C-terminal extension of the leader peptidase. Spontaneous insertion of H5 und H5EE into liposomes occurred only into leaky vesicles of the E. coli lipids. The membrane integrity was improved by the addition of an adequate amount of diacylglycerol (DAG) to the phospholipids. The leaky phospholipids were sealed by the addition of 3-4% DAG. The proteins H5 und H5EE showed a dependency of the membrane potential. Insertion occured more efficiently into YidC proteoliposomes when a stable membrane potential was generated. Proteoliposomes with reconstituted SecYEG translocase were also tested for protein insertion. Remarkedly, the protein M13 procoat H5EE efficiently inserted into SecYEG proteoliposomes, where the wildtype-like protein H5 did not.
Molekulare Dynamik der YidC-Membraninsertase aus Escherichia coli
(2011) Imhof, Nora; Kuhn, Andreas
The membrane insertase YidC of the Gram-negative bacterium E. coli enables the insertion of proteins into the cytoplasmic membrane. YidC itself is localized in the cytoplasmic membrane and spans the membrane six times with its N- and C-termini localized in the cytoplasm. These six transmembrane segments are connected by three periplasmic loops (P1, P2 and P3) and two cytoplasmic loops (C1 and C2). It is known that the binding of the YidC-dependent protein Pf3 coat induces conformational changes in the tertiary structure of YidC. This molecular dynamic of YidC was examined in detail with steady-state and time-resolved fluorescence spectroscopy. Therefore, three tryptophan mutants of YidC with one tryptophan residue each, at position 354 in the first periplasmic domain P1, at position 454 in the second periplasmic region and at position 508 near the third periplasmic region, respectively, were used. Additionally, a double tryptophan mutant was used which contained two tryptophan residues at position 332/334 of the domain P1. These tryptophan residues were used as intrinsic fluorophores. First, it was shown that the tryptophan mutants of YidC complemented the growth defect of the E. coli YidC-depletion strain JS7131. Additionally, the mutants were able to insert the strictly YidC-dependent PClep protein into the cytoplasmic membrane of the depletion strain. Thus, the functionality of the tryptophan mutants of YidC was ensured. Purified tryptophan mutants of YidC were reconstituted into liposomes and titrated with Pf3W0 coat, a tryptophan free mutant of Pf3 coat protein allowing spectroscopic studies of each periplasmic region (P1, P2 and P3) before and after binding of Pf3W0 coat protein. Analysis of the emission spectra and the fluorescence lifetimes of detergent solubilized as well as of the reconstituted YidC tryptophan mutants before binding of Pf3W0 coat revealed that the tryptophan residue of each single tryptophan mutant (YidCW354, YidCW454 and YidCW508) was localized at the membrane/water interface. These results are consistent with the proposed membrane topology of YidC. The tryptophan residues of the double tryptophan mutant of YidC (YidC2W) showed fluorescence properties consistent with their localization in a partially exposed alpha-helical segment of the P1 domain. Analysis of the emission spectra and the fluorescence lifetimes provided additional evidence that binding of Pf3W0 coat induced conformational changes of all periplasmic regions (P1, P2 and P3) within YidC. Measurements of fluorescence anisotropy showed that the conformational changes affected motions within all three periplasmic regions of the YidC tryptophan mutants, whereas the periplasmic domain P1 with the tryptophan residues W332/W334 and the third periplasmic domain P3 with the tryptophan residue W508 were affected most significantly.
The bacterial membrane insertase YidC : in vivo studies of substrate binding and membrane insertion
(2015) Klenner, Christian Daniel; Kuhn, Andreas
YidC of Escherichia coli belongs to the evolutionarily conserved proteins of the Oxa1/YidC/Alb3 insertase family. The transmembrane regions of the core domain, comprising of TM2-6, are the most conserved parts among the homologs and are crucial for the function as a membrane insertase. This is particularly true for the TM2, TM3 and TM5 (KUHN et al., 2003; KIEFER & KUHN, 2007). In bacteria, YidC acts as an independently working membrane insertase and, as well, in cooperation with the Sec translocon for the biogenesis of various membrane proteins. YidC is required for the biogenesis of respiratory complexes, ATP synthase and for example the mechanosensitive channel protein MscL. Also, the coat proteins of filamentous phage Pf3 and M13 require YidC for membrane insertion. The best studied substrate is the Pf3 coat protein of phage Pf3 infecting Pseudomonas aeruginosa – i.e. a small protein of 44 amino acids in length. In the context of this thesis, the YidC-dependent biogenesis of Pf3 coat was analyzed to gain better insight into the entire insertion process. In doing so, a set of more than 100 single cysteine mutants in distinct domains of YidC and Pf3 coat were generated. To study the insertion of Pf3 coat under physiological conditions, an in vivo cross-linking assay was established for capturing YidC-Pf3 interactions within a short period of time after the onset of synthesis (1 minute) using 35S-Met pulse-labelling methods. YidC binds inserting Pf3 coat protein in distinct regions of the highly conserved TM domains involving four of the six TM helices. It was verified that TM3 is indispensable for the function of YidC since four contacting residues were found in this TM helix. A helical wheel projection of substrate binding helices reveals the localization of the contacting residues of each TM segment on one helical face. This implies a helix arrangement of the transmembrane core domain which enables binding of inserting substrate proteins and interactions with transmembrane domains over the entire membrane-spanning part of YidC. The serial mutation of nine from twelve contacting residues, which are strongly hydrophobic in most cases, to serines impaired the function of YidC, whereas the single mutations had no effect. Additionally, the insertion process of translocation deficient Pf3 coat mutants was analyzed for intermediate states of the insertion process. It has been shown that the insertion deficient Pf3 coat mutants are inhibited at a late step of membrane insertion, i.e. forming the YidC contacts in the periplasmic leaflet. Based on this work, further studies confirmed that the identified substrate contacting regions of YidC play a key role in YidC-mediated insertion. The mechanosensitive channel protein MscL, M13 procoat, nascent Foc and the polytopic membrane protein LacY contact YidC at exactly the same positions (NEUGEBAUER et al., 2012; SPANN & KUHN, unpublished results; WICKLES et al., 2014; ZHU et al., 2013b).
The function of E. coli YidC for the membrane insertion of the M13 procoat protein
(2018) Spann, Dirk; Kuhn, Andreas
The YidC/Oxa1/Alb3 family consists of insertase homologues that facilitate the insertion and folding of membrane proteins. YidC is located in the inner membrane of bacterial cells. Oxa1 is found in the inner membrane of mitochondria and Alb3 facilitates the insertion of membrane proteins in the thylakoid membranes of chloroplasts (Wang and Dalbey 2011, Hennon et al. 2015). An archaeal homologue was found in M. jannaschii showing that this insertase family is present in all domains of life (Dalbey and Kuhn 2015). The insertase family shares a structural feature that is conserved among all discovered members. This is a hydrophilic groove that is open towards the cytoplasm and the membrane core with a hydrophobic slide formed by transmembrane domain (TM) 3 and TM5. YidC functions on its own but also cooperates with the Sec translocon to facilitate the insertion of large membrane proteins. One protein that is membrane-inserted by YidC but is Sec-independent is the major coat protein of the M13 bacteriophage. The main objectives of this work are the analysis of the insertion mechanism of M13 procoat, the major capsid protein of the M13 bacteriophage, via the YidC-only pathway and the oligomeric state of the active YidC. The analysis of interactions between YidC and M13 procoat was performed via radioactive disulfide crosslinking mainly using copper phenanthroline as oxidizing agent. M13 procoat contacts YidC extensively in TM3 and TM5. The observed contacts suggest that the M13 procoat substrate “slides” along TM3 and TM5 of the insertase. Additional crosslinking experiments with the hydrophilic groove and the C1 loop of YidC were also performed to test their importance during the insertion process. A contact was found in the C1 loop that indicates a role in the insertion process, which is consistent with the proposed insertion model from Kumazaki et al. (2014a). Parallel to the radioactive disulfide crosslinking, a protocol using DTNB (Bis(3-carboxy-4-nitrophenyl) disulfide, Ellman’s reagent) as the oxidizing reagent and Western blot for detection was established. This method reliably promoted the formation of crosslinking products in vivo between YidC and M13 procoat over several hours and many, but not all, mapped at the same sites as in the radioactive approach. In addition, this protocol was used to purify small amounts of a YidC-substrate complex for biochemical analysis, which could also be applied to other substrates in the future. The oligomeric state of YidC was investigated by an artificial dimer of the insertase (dYidC) that was constructed by connecting two monomers together with a short linker. This dimer can complement YidC-depleted E. coli MK6S cells and facilitates the insertion of M13 procoat in vivo. For further analysis of the dYidC three functionally defective YidC mutants, T362A (Wickles et al. 2014), delta-C1 (Chen et al. 2014) and the 5S YidC mutant, were tested for their complementation and insertion capability. All three mutants were not able to complement under YidC depletion conditions. These mutants were then cloned in either one or both protomers of the dYidC. Complementation and insertion assays with these dYidC constructs show that in general one active protomer suffices to uphold cell viability and to facilitate the insertion of M13 procoat. Binding studies using cysteine mutants of the dYidC and M13 procoat for disulfide crosslinking with DTNB demonstrated that each protomer individually binds one substrate molecule. In summary, these experiments strongly support a monomer as the active state of the insertase for YidC-only substrates. Taken together, this study contributes to the understanding of the insertion of proteins into the inner bacterial cell membrane.
Tth-IM60, eine Membraninsertase aus Thermus thermophilus
(2011) Meyer, Susanne H.; Kuhn, Andreas
The evolutionarily conserved YidC/Oxa1/Alb3 family of proteins catalyzes the insertion of integral membrane proteins in bacteria, mitochondria, and chloroplasts. In this work Tth-IM60 from Thermus thermophilus was identified as a member of this family. The function and structure of the protein was analysed in detail. Complementation studies in a Escherichia coli YidC-depletion strain show a functional replacement of the essential YidC by Tth-IM60 in vivo. A heterologous expression of the his-tagged Tth-IM60 protein was achieved in the E. coli strain C43 and pMS as a plasmid vector. It was shown that Tth-IM60 protein is located in the inner membrane probably in a dimeric state. After purification the protein tends to oligomerize in a higher, but very stable oligomeric state. The Tth-IM60 oligomeric protein was stable for 50 days at least. By size-exclusion chromatography the zwitterionic detergent LDAO and a buffer containing 20 mM TrisHCl pH 8,5, 500 mM NaCl and 10 % glycerol were identified as the best conditions for purification and stability. With this buffer, Tth-IM60 was purified as a dimer via its C-terminal histag by two Ni-IMACs (immobilized metal affinity chromatography) and a size-exclusion chromatography. A final protein concentration up to 10 mg/ml was feasible. The purified Tth-IM60 protein was used for functional and structural studies. A weak binding of Tth-IM60 to the first periplasmic loop of SecF as shown by pulldown assays, suggests a Sec-dependent function of Tth-IM60. In addition, the essential Sec-independent function of Tth-IM60 was demonstrated by the translocation of the YidC substrate Pf3 coat into Tth-IM60 proteoliposomes. Moreover, the results of the single molecule spectroscopy measurements implicate that the translocation of Pf3 coat proteins occurs with a fast kinetics within 5 minutes. The secondary structure of the Tth-IM60 protein was analysed by circular dichroism spectroscopy: 49 ? 55 % alpha-helical, 13 - 14 % beta-sheet, 14 ? 17 % beta-turns and 18 - 21% unordered structures were calculated. Tth-IM60 comprises more alpha-helical structures, but less beta-sheets than YidC, probably because of the first periplasmic loop of Tth-IM60 being shorter. The melting point of Tth-IM60 was determined to 68 °C, which is 10 degrees higher than the melting point of YidC. Furthermore, the Tth-IM60 protein was crystallized and X-ray analysis was performed. However, due to the low resolution the structure of the Tth-IM60 protein could not be determined so far. The second part of this thesis concerns the ?translocating chain-associated membrane? (TRAM) protein. The TRAM protein is involved in the insertion of integral membrane proteins of the endoplasmic reticulum (ER) together with the Sec61-complex. In contrast to the Sec-components, no homologous YidC protein exists in the ER-membrane. Therefore, it was postulated that TRAM and YidC could have functional similarities. For functional studies the TRAM protein from X. laevis was expressed in E. coli as a fusion protein with a N-terminal MBP (maltose-binding protein) and a C-terminal histag. TRAM was purified via two different affinity matrices: Ni sepharose and an amylose resin. It was possible to reconstitute the fusion protein into liposomes. However a translocation of the YidC-substrate Pf3 coat into these proteoliposomes was not detectable. In addition, complementation studies with a YidC-depletion strain did not show that the essential YidC function can be replaced by TRAM in vivo.
Untersuchungen zur autonomen und YidC-vermittelten Membraninsertion von Pf3 coat-Protein mit Hilfe Fluoreszenz-spektroskopischer Einzelmolekülmessungen
(2011) Schönbauer, Anne-Kathrin; Kuhn, Andreas
Pf3 coat is the capsid protein of the bacteriophage Pf3. The phage leaves the host cell by continuous extrusion without damaging the cell. The protein itself consists of 44 amino acid residues and has a rod-like shape. Because of its simple structure, the protein needs only the help of the insertase YidC to insert into the bacterial inner membrane. 3L-Pf3 coat, a protein mutant with three additional leucine residues in the center of the transmembrane region (TMD), has an increased hydrophobicity. It is independent of YidC and inserts into the membrane autonomously (Serek et al., 2004). In this work, a newly developed physical method was used to find out whether the elongation or the increased hydrophobicity accounts for the autonomous insertion of the protein. For this reason, two new protein mutants were constructed. Each mutant has only one of the changed properties of the 3L-Pf3 coat protein: GAT-Pf3 coat has an elongated TMD with three additional residues (glycine, alanin and threonine). The second mutant, 2M-Pf3 coat, shows an increased hydrophobicity due to the substitution of two alanine residues by two methionine residues at the positions 30 and 31. So it had an increased hydrophobicity like 3L-Pf3 coat. The above mentioned proteins, wt-Pf3 coat and its mutants, were modified with a fluorescent label to follow the proteins with optical methods. The Proteins were first modified with a single cysteine and then labeled by a fluorescent marker, Atto520 maleimid. Proteins with a labeled N-terminal tail were called NC-Pf3 coat, whereas CC-Pf3 coat had a labeled C-terminal tail. In addition, the orientation of the protein in the membrane was identified by quenching the fluorescence of the NC- and CC- labeled proteins. A new method employing single molecules was developed using fluorescence correlation spectroscopy. This method allows real time observations of binding and insertion of the protein into semisynthetical liposomes. By using fluorescent quenching the membrane insertion and binding were distinguished. It became clear that both the elongation of the TMD as well as an increased hydrophobicity play a crucial role in the autonomous insertion of the protein into the membrane. Therefore, the interaction between the hydrophobic region of the protein and the hydrophobic core region of the membrane is important for the binding of the protein and its insertion into the membrane.