In mono- und eudikotylen Pflanzen kodiert eine Genfamilie (RpoT, RNA-Polymerase des T3/T7-Typs) mitochondriale und plastidäre RNA-Polymerasen (RNAP), die den ungeraden T-Phagen-Polymerasen ähneln. RpoT-Gene von Angiospermen sind gut charakterisiert, während aus tiefer abzweigenden Pflanzenspecies bisher lediglich die Gene aus dem Moos Physcomitrella beschrieben wurden. Um einen Beitrag zur Aufklärung der molekularen Evolution der RpoT-Polymerasen im Pflanzenreich zu liefern und um Erkenntnisse über die potentielle Bedeutung von multiplen Phagen-Typ (RNAP) in Pflanzen zu gewinnen, wurden die RpoT-Gene aus dem Lycophyten Selaginella moellendorffii und aus dem basalen Angiosperm Nuphar advena identifiziert und charakterisiert. Selaginella moellendorffii (Moosfarn)-Trace-Sequenzdaten mit hoher Ähnlichkeit zu RpoT-Sequenzen von Angiospermen wurden benutzt, um das full-length SmRpoT-Gen und die entsprechende cDNA zu isolieren. Die SmRpoT-mRNA ist 3542 nt lang und weist einen offenen Leserahmen von 3006 nt auf, der für ein putatives Protein aus 1002 Aminosäuren mit einer molekularen Masse von 113 kDa kodiert. Das SmRpoT-Gen besteht aus 19 Exons und 18 Introns, die in ihren Positionen mit denen aus den Angiosperm- und Physcomitrella-Genen konserviert sind. Mittels Southernblot-Analyse wurde nachgewiesen, dass S. moellendorffii ein single-copy RpoT-Gen kodiert. Für das N-terminale Transitpeptid von SmRpoT konnte gezeigt werden, dass es bei transienter Expression in Arabidopsis- und Selaginella-Protoplasten den Transport von GFP (green fluorescent protein) exclusiv in Mitochondrien vermittelt. In N. advena wurden mittels Screening einer BAC-Bibliothek drei RpoT-Gene identifiziert. Sowohl die genomischen als auch die cDNA-Sequenzen wurden aufgeklärt. Die NaRpoT-mRNAs kodieren putative Polypeptide von 996, 990 und 985 Aminosären. Alle drei Gene besitzen 19 Exons und 18 Introns, die in ihren Positionen mit denen der RpoT-Gene aus Selaginella und allen anderen Landpflanzen konserviert sind. Die kodierten Proteine weisen auf Aminosäureebene einen hohen Konservierungsgrad auf, einschließlich aller essentiellen Regionen und Aminosäurereste, die für die T7-RNAP bekannt sind. Die N-terminalen Transitpeptide zweier der kodierten RNAP, NaRpoTm1 und NaRpoTm2, vermittelten den Import von GFP exclusiv in Mitochondrien, während die dritte Polymerase, NaRpoTp, in Chloroplasten importiert wurde. Interessanterweise muß die Translation der NaRpoTp-mRNA an einem CUG-Codon initiiert werden, um ein funktionelles Protein mit plastidärem Transitpeptid zu erhalten. Die N. advena RpoTp-RNAP ist somit neben AGAMOUS aus Arabidopsis und der RpoTp-RNAP aus Nicotiana, ein weiteres Beispiel für jene selten vorkommenden pflanzlichen mRNAs, deren Translation exclusiv an nicht-AUG-Codons initiiert wird. Die Rekonstruktion von phylogenetischen Bäumen resultierte in unterschiedlichen Positionen für die Selaginella- und Nuphar-Polymerasen: Im Gegensatz zu der RpoT-Polymerase aus S. moellendorffii und denen aus Physcomitrella, die in den phylogenetischen Analysen Schwesterpositionen zu allen anderen Phagentyp-RNAP der Angiospermen einnehmen, clusterten die Nuphar-RpoTs zusammen mit den deutlich separierten mitochondrialen (NaRpoTm1 und NaRpoTm2) und plastidären (NaRpoTp) Polymerasen. Selaginella kodiert eine einzige mitochondriale RNAP, während Nuphar zwei mitochondriale und eine plastidäre RNAP besitzt. Die Identifizierung einer Plastiden-lokalisierten Phagentyp-RNAP in diesem basalen Eudikotylen, die ortholog zu allen anderen RpoT-Enzymen der Blütenpflanzen ist, läßt darauf schließen, daß die Acquisition einer nukleär kodierten plastidären RNAP, die noch in den Lycopoden fehlt, nach der Trennung der Leucopoden von allen anderen Tracheophyten erfolgte. Eine “dual-targeting” RNAP (mitochondrial und plastidär lokalisiert), wie sie in Eudikotylen, nicht jedoch in Monokotylen vorkommt, wurde weder in Selaginella noch in Nuphar nachgewiesen, vermutlich ist sie ein evolutionäres Novum von eudikotylen Pflanzen wie Arabidopsis. / In mono- and eudicot plants, a small nuclear gene family (RpoT, RNA polymerase of the T3/T7 type) encodes mitochondrial as well as chloroplast RNA polymerases homologous to the T-odd bacteriophage enzymes. RpoT genes from angiosperms are well characterized, whereas data from deeper branching plant species until recently were limited to the moss Physcomitrella. To elucidate the molecular evolution of the RpoT polymerases in the plant kingdom and to get more insight into the potential importance of having more than one phage-type RNA polymerase (RNAP) available, we identified and characterized RpoT genes in the lycophyte Selaginella moellendorffii and the basal eudicot Nuphar advena. Selaginella moellendorffii (spikemoss) sequence trace data encoding a polypeptide highly similar to angiosperm and moss phage-type organelle RNA polymerases were used to isolate a BAC clone containing the full-length gene SmRpoT as well as the corresponding cDNA. The SmRpoT mRNA comprises 3452 nt with an open reading frame of 3,006 nt, encoding a putative protein of 1,002 amino acids with a molecular mass of 113 kDa. The SmRpoT gene comprises 19 exons and 18 introns, conserved in their position with those of the angiosperm and Physcomitrella RpoT genes. Using Southern blot analysis, it was shown that S. moellendorffii encodes a single RpoT gene. The N-terminal transit peptide of SmRpoT was shown to confer targeting of green fluorescent protein (GFP) exclusively to mitochondria after transient expression in Arabidopsis and Selaginella protoplasts. In Nuphar advena three RpoT genes were identified by BAC library screening. Both genomic gene sequences and full-length cDNAs were determined. The NaRpoT mRNAs specify putative polypeptides of 996, 990 and 985 amino acids, respectively. All three genes comprise 19 exons and 18 introns, conserved in their positions with those from S. moellendorffii and the RpoT genes of other land plants. The encoded proteins show a high degree of conservation at the amino acid sequence level, including all functional crucial regions and residues known from the phage T7 RNAP. The N-terminal transit peptides of two of the encoded polymerases, NaRpoTm1 and NaRpoTm2, conferred targeting of GFP exclusively to mitochondria, whereas the third polymerase, NaRpoTp, was targeted to chloroplasts. Remarkably, translation of NaRpoTp mRNA has to be initiated at a CUG codon to generate a functional plastid transit peptide. Thus, besides AGAMOUS in Arabidopsis and the Nicotiana RpoTp polymerase, N. advena RpoTp provides another example for a plant mRNA that is exclusively translated from a non-AUG codon. Reconstruction of phylogenetic trees revealed different positions of the RpoTs from the lycophyte Selaginella and the basal eudicot Nuphar. In contrast to the RpoTs of S. moellendorffii and those of the moss Physcomitrella, which are according to the phylogenetic analyses in sister positions to all other phage-type polymerases of angiosperms, the Nuphar RpoTs clustered with the well separated clades of mitochondrial (NaRpoTm1 and NaRpoTm2) and plastid (NaRpoTp) polymerases. Selaginella encodes a single mitochondrial RNAP, whereas Nuphar harbors two mitochondrial and one plastid phage-type polymerases. Identification of a plastid localized phage-type RNAP in this basal eudicot, orthologous to all other RpoTp enzymes of flowering plants, suggests that the acquisition of a nuclear encoded plastid RNA polymerase, not present in lycopods, took place after the split of lycopods from all other tracheophytes. A dual-targeted mitochondrial and plastid RNA polymerase (RpoTmp), as present in eudicots but not monocots, was not detected in Nuphar or Selaginella suggesting that its occurrence is an evolutionary novelty of eudicotyledoneous plants like Arabidopsis.

Lorsqu’il est surexprimé ou activé constitutivement par mutation, le récepteur tyrosine kinase KIT est impliqué dans le développement de pathologies prolifératives comme les mastocytoses, les tumeurs stromales gastro-intestinales (GIST) et certaines leucémies. La voie de signalisation KIT représente donc une cible thérapeutique majeure en oncologie. Le développement d’une nouvelle classe de molécules pharmacologiques appelées inhibiteurs de tyrosine kinase (ITK) est en plein essor. Un exemple majeur d’ITK est l’imatinib qui cible, entre autre, KIT et est efficace dans la plupart des GIST. Cependant, le traitement aux ITK est souvent confronté au phénomène de résistance primaire ou acquise par mutation secondaire. C’est pourquoi nous cherchons à développer de nouveaux composés ciblant KIT ou les voies de transductions activées par ses formes oncogéniques, et ce par 3 approches.Nous avons récemment montré que les mutants oncogéniques de KIT ont une localisation intracellulaire alors que KIT sauvage est exprimé à la membrane. L’inhibition de l’activité kinase des mutants restaure une localisation normale. A partir de cette observation, nous avons créé et validé un test de criblage par cytométrie mesurant la relocalisation de KIT muté à la surface cellulaire. Le criblage d’une chimiothèque nous a permis de sélectionner de nouveaux inhibiteurs de la signalisation KIT actifs sur des lignées cellulaires mutées pour KIT.Nous avons utilisé la technique du phage display pour sélectionner des anticorps au format scFv et VHH spécifiques de la partie intracellulaire de KIT mutant. Lors de leur expression dans le cytosol (on parle alors d’intrabodies), leur fixation au niveau de KIT inhibe soit directement l’activité kinase, soit le recrutement de partenaires de signalisation. Nous avons obtenu des intrabodies de différentes spécificités vis-à-vis des formes de KIT dont la caractérisation fonctionnelle est en cours Les intrabodies inhibiteurs seront utilisés pour cribler des chimiothèques par ELISA. Les molécules chimiques recherchées empêcheront la fixation des intrabodies sur la région intracellulaire de KIT. On sélectionnera donc des molécules inhibant potentiellement l’oncogénicité de KIT.Nous avons développé des anticorps au format scFv-Fc par phage display qui reconnaissent le domaine extracellulaire de KIT. Deux des anticorps sélectionnés inhibent donc la signalisation induite par le SCF. Dans des lignées de leucémie exprimant KIT WT, nous avons montré que l’utilisation de ces anticorps entraîne une diminution de la viabilité cellulaire. De plus, ils diminuent également la prolifération de lignées de leucémie à mastocytes sensibles et résistantes à l’imatinib (HMC11 et HMC12, respectivement). Ils représentent donc des outils thérapeutiques potentiels pour le traitement des pathologies impliquant KIT ainsi que pour contourner la résistance aux ITK de certains mutants. / When overexpressed or constitutively active by mutation, the tyrosine kinase receptor KIT is involved in some proliferative diseases such as gastro-intestinal stromal tumors (GIST), mastocytosis and some leukemia. Therefore, KIT signaling represents a major target in oncology. Development of a new therapeutic class called tyrosine kinase inhibitors is in full expansion. A major example of TKI is imatinib which targets KIT and is efficient in the majority of GIST cases. However, TKI treatment is often unpaired by primary or acquired resistance due to secondary mutations. That is why we aim to develop new compounds to target KIT or associated signaling pathways by three strategies.We have recently shown that oncogenic KIT mutants are intracellularely localized whereas WT KIT is expressed at the cell surface. Kinase activity inhibition leads to membrane mutants’ relocalization. Based on this finding, we developed and validated a screening assay measuring mutants’ relocalization by cytometry. Chemicals library screening allows us to select new KIT signaling inhibitors active on KIT mutant cell lines.We used phage display to generate scFv and VHH antibodies which are specific to KIT intracellular domain. When expressed in cytoplasm (they are called intrabodies), their binding on KIT inhibits kinase activity directly or signaling partners’ recruitment. Selected intrabodies are specific to various KIT isoforms and their functional characterization is ongoing. KIT inhibitory intrabodies will be used to screen chemical libraries by ELISA for drugs that block intrabodies binding on KIT intracellular domain. We will then select molecules that potentially inhibit KIT oncogenicity.We developed scFv-Fc antibodies by phage display that recognize KIT extracellular domain. Two selected antibodies inhibit SCF induced signaling. In WT KIT expressing leukemic cell lines, we showed that antibody treatment reduces cell viability. Moreover, they also diminish cell proliferation of 2 imatinib sensitive and resistant mast cell leukemia cell lines (HMC1.1 and HMC1.2, respectively). They represent potential therapeutic tools for treatment of KIT involved diseases and for bypass TKI resistance of some mutants.

No description available.

No description available.

O uso de diferentes tipos de medicação intracanal para o controle do processo infeccioso, principalmente nos casos em que há presença de microrganismos resistentes às manobras de desinfecção, tem sido alvo de muitas pesquisas. A proposta deste estudo foi avaliar, in vitro, o efeito antimicrobiano de bacteriófagos específicos diante de cinco cepas de Enterococcus faecalis e a ação de um lisado híbrido polivalente na eliminação da infecção causada por essas cinco cepas da mesma espécie. Foram utilizados 37 dentes unirradiculares humanos, recentemente extraídos e de proporções aproximadas. As coroas foram removidas e os canais instrumentados até a lima tipo K de número 45. Os espécimes foram, então, esterilizados e utilizados em dois experimentos distintos. O primeiro experimento utilizou 25 raízes divididas em cinco grupos de cinco espécimes. Três espécimes de cada grupo foram inoculados com uma das culturas bacterianas e seus fagos correspondentes na proporção 1:1, por um período de três horas a 37 °C, enquanto os outros dois, receberam a cultura de microrganismos ou somente meio de cultura (controle positivo e negativo, respectivamente). No segundo experimento, 11 espécimes receberam um inóculo formado pelas cinco cepas por um período de 10 dias de incubação a 37 °C, com o propósito de manter condições apropriadas para a penetração das bactérias no interior dos túbulos dentinários, e um outro espécime recebeu apenas meio de cultura (controle negativo). Essa penetração foi confirmada empregando-se microscopia ótica e eletrônica realizada em dois espécimes. Após o período de incubação, o lisado polivalente, preparado com os cinco fagos, foi aplicado por 24 horas a 37 °C em 8 espécimes, e os demais preenchidos com meio de cultura (controle positivo e negativo). Alíquotas do interior de todos os canais foram colhidas antes e depois do contato com os fagos e no segundo experimento, também 24 e 48 horas depois, para semeadura e contagem de unidades formadoras de colônia. Os resultados do primeiro experimento mostraram 100% de redução do crescimento bacteriano nos espécimes que receberam a suspensão de fagos específicos, em comparação a seus respectivos controles positivos, em todos os grupos. No segundo experimento, foi comparado o crescimento obtido após os 10 dias de infecção com aquele posterior a aplicação dos fagos, redução que variou entre 50% e 100%. Diante desses resultados, conclui-se que os bacteriófagos foram eficazes na diminuição dos microrganismos presentes no interior de canais radiculares e nos túbulos dentinários de dentes humanos. / Many studies have investigated different intracanal medications to control infection processes, especially in cases of microbial resistance to disinfection procedures. The purpose of this study was to evaluate the in vitro antimicrobial effect of specific bacteriophages on five isolates of Enterococcus faecalis, as well as the activity of a lysate cocktail in eliminating the infection caused by these bacteria. Thirty-seven recently extracted human teeth of approximately equal size and with single roots were used. The crowns were removed and each canal was prepared using K files,up to # 45, and sterile physiological saline. Specimens were then sterilized and used in two separate studies. The first study utilized 25 individual roots divided into five groups of five specimens each. Three specimens of each group were inoculated with one of the bacterial cultures and the corresponding bacteriophage in a proportion of 1:1, and incubated for three hours at 37°C; the other two specimens were inoculated with only the bacterial culture or only the culture medium (positive and negative controls, respectively). In the second study, 11 specimens were inoculated with all five strains and incubated for ten days at 37°C in order to allow bacteria to penetrate the interior of the dental tubules, and another one, received just the culture medium (negative control). Penetration into the tubules was confirmed by optical and electron microscopy of two specimens. Following incubation, the lysate cocktail prepared using all five bacteriophages was applied to the other 8 specimens for 24 hours at 37°C, and 2 specimens were filled with the culture medium (positive and negative controls). In the first study, samples were taken from the lumen of all canals before and after contact with bacteriophages; in the second, aliquots were also taken 24 and 48 hours after the bacteria were exposed to the phages. All samples were diluted and plated and the number of colony forming units was counted. In the first study, there was a 100% reduction in bacterial growth in specimens that received the specific bacteriophage suspension compared to the positive controls within each group. In the second study, after ten days the number of bacteria was reduced by 50% to 100% following the bacteriophage application. These results suggest that bacteriophages are effective in reducing the number of bacteria inside the root canal and in the dental tubules of human teeth.

<div> <div> <div> <p>Mycobacteriophages are the viruses that infect mycobacteria. Due to the high death rate and antibiotic-resistant strains, phage therapy is considered to be a promising treatment of tuberculosis. Current understanding of phage-bacteria interaction is abstracted as phage lytic and lysogenic life cycles. However, bacterial physiology may impact phage life cycles and bacterial cells with different physiology may have different responses to phage infection. In order to improve the understanding of phage-bacteria interaction and update phage therapy strategy, the impact of mycobacterial physiology on mycobacteriophage life cycles was studied in this research. In this research, a mass spectrometry-based method was first developed to study phage proteins in phage-bacteria mixture. Then five mycobacteriophages isolated at Purdue University were selected to infect exponential and stationary <i>Mycobacterium smegmatis</i> (<i>M. smegmatis</i>) cell cultures. Growth curves of the <i>M. smegmatis</i> cell cultures infected by the five phages were determined. Proteomics and lipidomics of the <i>M. smegmatis</i> cells cultures infected by phages FrenchFry and MrGordo were analyzed by mass spectrometry. The correlations between individual proteins/lipids and the experimental factors (bacterial growth phases, phages and phage infection time) were studied by developing linear regression models using SAS. The mass spectrometry-based method was proved to be able to detect phage proteins other than the structural proteins. It also verified the phage protein annotation that had been accomplished <i>in silico</i>. X! Tandem and a database consisting of six frame translation of the phage genome and the annotated proteins of <i>M. smegmatis</i> were the optimal option for analyzing mass spectra data of phage-bacteria mixture. The growth curves of the <i>M. smegmatis</i> infected by the phages displayed that growth of exponential <i>M. smegmatis</i> cell cultures were depressed by phages (except FrenchFry) and stationary <i>M. smegmatis</i> cell cultures were not actively lysed by any of the phages. The proteomics results showed that MrGrodo infection impacted more proteins than other factors did. Exponential phase up-regulated proteins involved in cell division. Stationary phase up-regulated proteins that may change cell surface properties. FrenchFry up-regulated LuxR protein. Infection time up-regulated the proteins associated with mycobacterial virulence. The lipidomics results indicated that growth phases impacted the most lipids. Phage infection time increased the amount of the lipids related to mycobacterial virulence. In summary, the mass spectrometry-based method developed in this research can be employed to study phage proteins in phage-bacteria mixture and verify phage genome annotation. Mycobacterial physiology alters mycobacteriophage life cycles. Phage-bacteria interaction is the interaction between the two populations instead of between an individual phage particle and an individual bacterial cell. Virulence of <i>M. smegmatis</i> improves as a response to phage infection.</p></div></div></div>

Mon travail de thèse a porté sur le ciblage de la vascularisation tumorale : des molécules ciblant l'endothélium tumoral ont été recherchées dans le but d'amener des molécules thérapeutiques spécifiquement vers l'endothélium tumoral afin de le détruire et d'atteindre la tumeur qui en dépendait. D'après les données de la littérature, deux cibles ont été choisies pour notre étude : la sélectine E et l'endogline. La sélectine E est une glycoprotéine surexprimée à la surface de l'endothélium activé des zones inflammatoires et tumorales. Classiquement, des antagonistes de son ligand naturel, le SleX, sont recherchés. En nous appuyant sur les groupements clés de l'interaction entre la sélectine E et le SleX, plusieurs familles de mimes du SleX ont été élaborées. La capacité de ces mimes à déplacer l'interaction du SleX exprimé à la surface des cellules HL-60 avec la sélectine E a été évaluée dans un test d'adhésion. Cependant, aucun des mimes testés n'a présenté une affinité suffisante pour envisager son utilisation comme tête de ciblage. Une deuxième stratégie a consisté à rechercher des ligands peptidiques de la sélectine E en criblant des HUVECs activées avec une banque de peptides sur phages. Plusieurs phages-peptides testés en ELISA ont montré une meilleure affinité pour la sélectine E et / ou avec la sélectine P par rapport au phage sans insert. Des tests de compétition avec des peptides synthétiques correspondants permettront d'évaluer leur spécificité pour la cible. En ce qui concerne l'endogline, des ligands peptidiques ont été recherchés avec une banque de peptides sur phages. Dans un premier temps, le gène codant l'endogline humaine a été cloné dans un vecteur d'expression eucaryote afin de réaliser la sélection sur la protéine cellulaire. Une sélection sur la protéine recombinante a été réalisée par la suite pour diminuer le bruit de fond lié aux cellules. Parmi les peptides obtenus, certains ont montré des homologies de séquence avec des ligands de l'endogline et le test de ces phages-peptides en ELISA sur la protéine recombinante a donné un fort signal par rapport au phage sans insert. Ces peptides seront caractérisés prochainement. En conclusion, des ligands peptidiques des sélectines E et P et de l'endogline ont peut-être été identifiés. Ce travail a par ailleurs permis de mettre en place les outils nécessaires à l'utilisation de la technologie de sélection avec des banques de peptides sur phages dans les meilleures conditions possibles.

Protein engineering and in vitro selection systems are powerful methods to generate binding proteins. In nature, antibodies are the primary affinity proteins and their usefulness has led to a widespread use both in basic and applied research. By means of combinatorial protein engineering and protein library technology, smaller antibody fragments or alternative non-immunoglobulin protein scaffolds can be engineered for various functions based on molecular recognition. In this thesis, a 46 amino acid small albumin-binding domain derived from streptococcal protein G was evaluated as a scaffold for the generation of affinity proteins. Using protein engineering, the albumin binding has been complemented with a new binding interface localized to the opposite surface of this three-helical bundle domain. By using in vitro selection from a combinatorial library, bispecific protein domains with ability to recognize several different target proteins were generated. In paper I, a bispecific albumin-binding domain was selected by phage display and utilized as a purification tag for highly efficient affinity purification of fusion proteins. The results in paper II show how protein engineering, in vitro display and multi-parameter fluorescence-activated cell sorting can be used to accomplish the challenging task of incorporating two high affinity binding-sites, for albumin and tumor necrosis factor-alpha, into this new bispecific protein scaffold. Moreover, the native ability of this domain to bind serum albumin provides a useful characteristic that can be used to extend the plasma half-lives of proteins fused to it or potentially of the domain itself. When combined with a second targeting ability, a new molecular format with potential use in therapeutic applications is provided. The engineered binding proteins generated against the epidermal growth factor receptors 2 and 3 in papers III and IV are aimed in this direction. Over-expression of these receptors is associated with the development and progression of various cancers, and both are well-validated targets for therapy. Small bispecific binding proteins based on the albumin-binding domain could potentially contribute to this field. The new alternative protein scaffold described in this thesis is one of the smallest structured affinity proteins reported. The bispecific nature, with an inherent ability of the same domain to bind to serum albumin, is unique for this scaffold. These non-immunoglobulin binding proteins may provide several advantages as compared to antibodies in several applications, particularly when a small size and an extended half-life are of key importance. / <p>QC 20121122</p>

The family 1 cellulose-binding modules (CBM1) form a groupof small, stable carbohydrate-binding proteins. These modulesare essential for fungal cellulosedegradation. This thesisdescribes both functional studies of the CBM1s as well asprotein engineering of the modules for several objectives. The characteristics and specificity of CBM1s from theTrichoderma reeseiCel7A and Cel6A, along with severalother wild type and mutated CBMs, were studied using bindingexperiments and transmission electron microscopy (TEM). Datafrom the binding studies confirmed that the presence of onetryptophan residue on the CBM1 binding face enhances itsbinding to crystalline cellulose. The twoT. reeseiCBM1s as well as the CBM3 from theClostridium thermocellumCipA were investigated by TEMexperiments. All three CBMs were found to bind in lineararrangements along the sides of the fibrils. Further analysesof the bound CBMs indicated that the CBMs bind to the exposedhydrophobic surfaces, the so called (200) crystalline face ofValoniacellulose crystals. The function and specificity of CBM1s as a part of an intactenzyme were studied by replacing the CBM from the exo-actingCel7A by the CBM1 from the endoglucanase Cel7B. Apart fromslightly improved affinity of the hybrid enzyme, the moduleexchange did not significantly influence the function of theCel7A. This indicates that the two CBM1s are analogous in theirbinding properties and function during cellulosehydrolysis. The CBM1 was also used for immobilization studies. Toimprove heterologous expression of a CBM1-lipase fusionprotein, a linker stability study was carried out inPichia pastoris. A proline/threonine rich linker peptidewas found to be stable for protein production in this host. Forwhole bacterial cell immobilization, theT. reeseiCel6A CBM1 was expressed on the surface of thegram-positive bacteria,Staphylococcus carnosus. The engineeredS. carnosuscells were shown to bind cellulosefibers. To exploit the stable CBM1 fold as a starting point forgenerating novel binders, a phage display library wasconstructed. Binding proteins against an amylase as well asagainst a metal ion were selected from the library. Theamylase-binding proteins were found to bind and inhibit thetarget enzyme. The metal binding proteins selected from thelibrary were cloned on the surface of theS. carnosusand clearly enhanced the metal bindingability of the engineered bacteria. <b>Keywords</b>: cellulose-binding, family 1carbohydrate-binding module, phage display, bacterial surfacedisplay, combinatorial protein library, metal binding, proteinengineering,Trichoderma reesei, Staphyloccus carnosus.

Carbohydrate recognition is a ubiquitous principle underlying many fundamental biological processes like fertilization, embryogenesis and viral infections. But how carbohydrate specificity and affinity induce a molecular event is not well understood. One of these examples is bacteriophage P22 that binds and infects three distinct Salmonella enterica (S.) hosts. It recognizes and depolymerizes repetitive carbohydrate structures of O antigen in its host´s outer membrane lipopolysaccharide molecule. This is mediated by tailspikes, mainly β helical appendages on phage P22 short non contractile tail apparatus (podovirus). The O antigen of all three Salmonella enterica hosts is built from tetrasaccharide repeating units consisting of an identical main chain with a distinguished 3,6 dideoxyhexose substituent that is crucial for P22 tailspike recognition: tyvelose in S. Enteritidis, abequose in S. Typhimurium and paratose in S. Paratyphi. In the first study the complexes of P22 tailspike with its host’s O antigen octasaccharide were characterized. S. Paratyphi octasaccharide binds less tightly (ΔΔG≈7 kJ/mol) to the tailspike than the other two hosts. Crystal structure analysis of P22 tailspike co crystallized with S. Paratyphi octasaccharides revealed different interactions than those observed before in tailspike complexes with S. Enteritidis and S. Typhimurium octasaccharides. These different interactions occur due to a structural rearrangement in the S. Paratyphi octasaccharide. It results in an unfavorable glycosidic bond Φ/Ψ angle combination that also had occurred when the S. Paratyphi octasaccharide conformation was analyzed in an aprotic environment. Contributions of individual protein surface contacts to binding affinity were analyzed showing that conserved structural waters mediate specific recognition of all three different Salmonella host O antigens. Although different O antigen structures possess distinct binding behavior on the tailspike surface, all are recognized and infected by phage P22. Hence, in a second study, binding measurements revealed that multivalent O antigen was able to bind with high avidity to P22 tailspike. Dissociation rates of the polymer were three times slower than for an octasaccharide fragment pointing towards high affinity for O antigen polysaccharide. Furthermore, when phage P22 was incubated with lipopolysaccharide aggregates before plating on S. Typhimurium cells, P22 infectivity became significantly reduced. Therefore, in a third study, the function of carbohydrate recognition on the infection process was characterized. It was shown that large S. Typhimurium lipopolysaccharide aggregates triggered DNA release from the phage capsid in vitro. This provides evidence that phage P22 does not use a second receptor on the Salmonella surface for infection. P22 tailspike binding and cleavage activity modulate DNA egress from the phage capsid. DNA release occurred more slowly when the phage possessed mutant tailspikes with less hydrolytic activity and was not induced if lipopolysaccharides contained tailspike shortened O antigen polymer. Furthermore, the onset of DNA release was delayed by tailspikes with reduced binding affinity. The results suggest a model for P22 infection induced by carbohydrate recognition: tailspikes position the phage on Salmonella enterica and their hydrolytic activity forces a central structural protein of the phage assembly, the plug protein, onto the host´s membrane surface. Upon membrane contact, a conformational change has to occur in the assembly to eject DNA and pilot proteins from the phage to establish infection. Earlier studies had investigated DNA ejection in vitro solely for viruses with long non contractile tails (siphovirus) recognizing protein receptors. Podovirus P22 in this work was therefore the first example for a short tailed phage with an LPS recognition organelle that can trigger DNA ejection in vitro. However, O antigen binding and cleaving tailspikes are widely distributed in the phage biosphere, for example in siphovirus 9NA. Crystal structure analysis of 9NA tailspike revealed a complete similar fold to P22 tailspike although they only share 36 % sequence identity. Moreover, 9NA tailspike possesses similar enzyme activity towards S. Typhimurium O antigen within conserved amino acids. These are responsible for a DNA ejection process from siphovirus 9NA triggered by lipopolysaccharide aggregates. 9NA expelled its DNA 30 times faster than podovirus P22 although the associated conformational change is controlled with a similar high activation barrier. The difference in DNA ejection velocity mirrors different tail morphologies and their efficiency to translate a carbohydrate recognition signal into action. / Kohlenhydraterkennung ist ein fundamentales Prinzip vieler biologischer Prozesse wie z.B. Befruchtung, Embryogenese und virale Infektionen. Wie aber Kohlenhydratspezifität und –affinität in ein molekulares Ereignis übersetzt werden, ist nicht genau verstanden. Ein Beispiel für ein solches Ereignis ist die Infektion des Bakteriophage P22, der drei verschiedene Salmonella enterica (S.) Wirte besitzt. Er erkennt und depolymerisiert die repetitiven Einheiten des O Antigens im Lipopolysaccharid, das sich in der äußeren Membran seines Wirtes befindet. Dieser Schritt wird durch die Tailspikes vermittelt, β helicale Bestandteile des kurzen, nicht kontraktilen Schwanzapparates von P22 (Podovirus). Das O Antigen aller drei Salmonella enterica Wirte besteht aus sich wiederholenden Tetrasacchariden. Sie enthalten die gleiche Hauptkette aber eine spezifische 3,6 Didesoxyhexose Seitenkette, die für die P22 Tailspikeerkennung essentiell ist: Tyvelose in S. Enteritidis, Abequose in S. Typhimurium und Paratose in S. Paratyphi. Im ersten Teil der Arbeit wurde die Komplexbildung von P22 Tailspike mit O Antigen Octasaccharidfragmenten der drei verschiedenen Wirte untersucht. S. Paratyphi Octasaccharide binden mit einer geringeren Affinität (ΔΔG≈7 kJ/mol) an den Tailspike als die beiden anderen Wirte. Die Kristallstrukturanalyse des S. Paratyphi Octasaccharides komplexiert mit P22 Tailspike offenbarten unterschiedliche Interkationen als vorher mit S. Enteritidis und S. Typhimurium Oktasaccharidkomplexen mit Tailspike beobachtet wurden. Diese unterschiedlichen Interaktionen beruhen auf einer strukturellen Änderung in den Φ/Ψ Winkeln der glykosidischen Bindung. Die Beiträge von verschiedenen Proteinoberflächenkontakten zur Affnität wurden untersucht und zeigten, dass konservierte Wasser in der Struktur die spezifische Erkennung aller drei Salmonella Wirte vermittelt. Obwohl die verschiedenen O Antigen Strukturen unterschiedliches Bindungsverhalten auf der Tailspikeoberfläche zeigen, werden alle vom Phagen P22 erkannt und infiziert. Daher wurde in einer zweiten Studie die multivalente Bindung zwischen P22 Tailspike und O Antigen charakterisiert. Die Dissoziationskonstanten des Polymers waren drei Mal langsamer als für das Oktasaccharid allein, was auf eine hohe Affinität des O Antigens schließen lässt. Zusätzlich wurde gezeigt, dass die Aggregate des Lipopolysaccharids in der Lage sind, die Infektiösität vom P22 Phagen zu reduzieren. Ausgehend davon wurde in einer dritten Studie die Bedeutung der Kohlenhydrat Erkennung auf den Infektionsprozess untersucht. Große S. Typhimurium Lipopolysaccharide Aggregate bewirkten die DNA Freisetzung vom P22 Kapsid. Dies deutet darauf, dass der P22 Phage keinen weiteren Rezeptor für die Infektion auf der Oberflächen seines Wirtes verwendet. Zusätzlich moduliert die P22 Tailspike Aktivität den Ausstoss der DNA vom P22 Phagen: Er ist langsamer, wenn der Phage Tailspikes besitzt, die weniger hydrolytisch aktiv sind und wurde nicht induziert, wenn Lipopolysaccharid eingesetzt wurde, dass zuvor mit Tailspike hydrolysiert wurde. Darüber hinaus wurde der Start der DNA Ejektion verzögert, wenn Tailspikes mit verminderter Affinität am Phagen vorhanden waren. Die Ergebnisse führten zu einem Modell für die Infektion von P22: Tailspikes positionieren den Phagen auf Salmonella enterica und ihre Aktivität drückt ein zentrales Strukturprotein des Phagen, das Stöpselprotein, auf die Membranoberfläche. Aufgrund des Membrankontaktes findet eine Konformationsänderung statt die zur Ejektion der Pilotproteine und zur Infektion führt. Vorhergehende Studien haben bisher nur die DNA Ejektion in vitro für Viren mit langen, nicht kontraktilen Schwänzen (Siphoviren) mit Proteinrezeptoren untersucht. In dieser Arbeit wurde das erste Mal die DNA Ejektion für einen Podovirus mit LPS Erkennung in vitro gezeigt. Die O Antigen Erkennung und Spaltung durch Tailspikeproteine gibt es häufig in der Phagenbiosphere, z.B. am Siphovirus 9NA. Die Kristallstrukturanalyse von 9NA Tailspike zeigt eine komplett gleiche Struktur, obwohl beide Proteine nur zu 36% Sequenzidentität besitzen. Zusätzlich hat 9NA Tailspike ähnliche enzymatische Eigenschaften. Diese ist für den DNA Ejektionsprozess im Siphovirus 9NA verantwortlich, der auch durch LPS Agreggate induziert wird. 9NA stößt dabei seine DNA 30 Mal schneller aus als Podovirus P22 obwohl die damit verbundene Konformationsänderung mit einer ähnlich hohen Aktivierungsbarriere kontrolliert wird. Daher spiegeln die Unterschiede in der DNA Ejektionsgeschwindigkeit der verschiedenen Tailmorphologien die Effezienz wieder, mit der die spezifische Kohlenhydraterkennung in ein Signal umgewandelt wird.