Many applications in biomedicine, research, and industry require recognition agents with specificity and selectivity for their target. Protein engineering enables the design of scaffolds that can bind targets of interest while increasing their stability, and expanding the scope of applications in which these scaffolds will be useful. Repeat proteins are instrumental in a wide variety of biological processes, including the recognition of pathogen-associated molecular patterns by the immune system. A number of successes using alternative immune system repeat protein scaffolds have expanded the scope of recognition agents available for targeting glycans and glycoproteins in particular. We have analyzed the innate immune genes of a freshwater polyp and found that they contained particularly long contiguous domains with high sequence similarity between repeats in these domains. We undertook statistical design to create a binding protein based on the H. magnipapillata innate immune TPR proteins. My second research project focused on creating a protein to bind cellulose, as it is the most abundant and inexpensive source of biomass and therefore is widely considered a possible source for liquid fuel. However, processing costs have kept lignocellulosic fuels from competing commercially with starch-based biofuels. In recent years a strategy to protect processing enzymes with synergistic proteins emerged to reduce the amount of enzyme necessary for lignocellulosic biofuel production. Simultaneously, protein engineering approaches have been developed to optimize proteins for function and stability enabling the use of proteins under non-native conditions and the unique conditions required for any necessary application. We designed a consensus protein based on the carbohydrate-binding protein domain CBM1 that will bind to cellulosic materials. The resulting designed protein is a stable monomeric protein that binds to both microcrystalline cellulose and amorphous regenerated cellulose thin films. By studying small changes to the binding site, we can better understand how these proteins bind to different cellulose-based materials in nature and how to apply their use to industrial applications such as enhancing the saccharification of lignocellulosic feedstock for biofuel production. Biomaterials made from natural human hair keratin have mechanical and biochemical properties that make them ideal scaffolds for tissue engineering and wound healing. However, the extraction process leads to protein degradation and brings with it byproducts from hair, which can cause unfavorable immune responses. Recombinant keratin biomaterials are free from these disadvantages, while heterologous expression of these proteins allows us to manipulate the primary sequence. We endeavored to add an RGD sequence to facilitate cell adhesion to the recombinant keratin proteins, to demonstrate an example of useful sequence modification. / Doctor of Philosophy / Many applications in medicine and research require molecular sensors that bind their target tightly and selectively, even in complex mixtures. Mammalian antibodies are the best-studied examples of these sensors, but problems with the stability, expense, and selectivity of these antibodies have led to the development of alternatives. In the search for better sensors, repeat proteins have emerged as one promising class, as repeat proteins are relatively simple to design while being able to bind specifically and selectively to their targets. However, a drawback of commonly used designed repeat proteins is that their targets are typically restricted to proteins, while many targets of biomedical interest are sugars, such as those that are responsible for blood types. Repeat proteins from the immune system, on the other hand, bind targets of many different types. We looked at the unusual immune system of a freshwater polyp as inspiration to design a new repeat protein to recognize nonprotein targets. My second research project focused on binding cellulose, as it is the most abundant and inexpensive source of biological matter and therefore is widely considered a possible source for liquid fuel. However, processing costs have kept cellulose-based fuels from competing commercially with biofuel made from corn and other starchy plants. One strategy to lower costs relies on using helper proteins to reduce the amount of enzyme needed to break down the cellulose, as enzymes are the most expensive part of processing. We designed such a protein for this function to be more stable than natural proteins currently used. The resulting designed protein binds to multiple cellulose structures. Designing a protein from scratch also allows us to study small changes to the binding site, allowing us to better understand how these proteins bind to different cellulose-based materials in nature and how to apply their use to industrial applications. Biomaterials made from natural human hair keratin have mechanical and biochemical properties that make them ideal for tissue engineering and wound healing applications. However, the process by which these proteins are extracted from hair leads to some protein degradation and brings with it byproducts from hair, which can cause unfavorable immune responses. Making these proteins synthetically allows us to have pure starting material, and lets us add new features to the proteins, which translates into materials better tailored for their applications. We discuss here one example, in which we added a cell-binding motif to a keratin protein sequence.

Introdução e Objetivos. Fosfatases são enzimas promissoras para aplicação na degradação de organofosforados. Por exemplo, a enzima paraoxonase 1 (PON1), associada à lipoproteína de alta densidade (HDL), hidrolisa lactonas, ésteres aromáticos e compostos organofosforados (OP) neurotóxicos. \"Módulos de ligação a carboidrato\" (CBM) têm diversas aplicações biotecnológicas. Nosso objetivo é a obtenção de proteínas quiméricas contendo fosfatases ligadas a um módulo de ligação de celulose, o que possibilitaria a imobilização dessas enzimas em suportes de celulose. Resultados. Como prova de conceito, uma proteína quimérica contendo uma \"fosfatase ácida\" (appA) de E.coli e CBM familia 2 (CBM2) de uma celulase de Xanthomonas axonopodis pv citri foi montada e produzida em E. coli como uma proteína recombinante solúvel. appA-CBM2 purificada demonstrou ser totalmente funcional exibindo atividade de ligação à celulose microcristalina (Avicel PH101) e atividade de fosfatase sobre p-nitrofenil fosfato. A ligação à Avicel evidenciou um comportamento de saturação descrito por uma \"constante de ligação\" (Kb) de 26 mg e um \"máximo de ligação\" (Bmax) de 4,45 U/µg. Além disso, a ligação de appA-CBM2 em Avicel foi maior em pH 2,5 e diminuiu acima de pH 6,5, como observado anteriormente para CBM2. Finalmente, o efeito de concentração de p-nitrofenil fosfato na atividade catalítica de appA-CBM2 e appA foi idêntico, exibindo um Km de 2,8 mM. Portanto, esses dados mostram que o conceito de uma proteína que combina as propriedades da fosfatase e do domínio de ligação à celulose é possível e funcional. De forma similar, os segmentos de DNA que codificam para o CBM2 e para a PON1 de Homo sapiens, foram fusionados resultando em um segmento que codifica para uma proteína quimérica (PON1-CBM2). PON1 nativa e PON1-CBM2 foram produzidas na forma solúvel e ativa em E.coli cepa Arctic. Embora não tenha sido viável sua purificação, estas enzimas foram caracterizadas. PON1-CBM2 liga-se em Avicel PH101 com um comportamento de saturação, descrito por uma constante de ligação (Kb) de 27 mg, valor idêntico àquele observado para appA-CBM2, o que sugere que o domínio CBM2 é igualmente funcional nestas duas enzimas quiméricas. PON1-CBM2 também exibe atividade paraoxonásica com Km similar àquele observado para PON1 nativa (1,3 mM), sugerindo que o \"domínio\" PON1 encontra-se totalmente funcional na enzima quimérica. Conclusão. Uma estratégia para a construção e expressão heteróloga em E. coli de PON1 e das enzimas quiméricas appA-CBM2 e PON1-CBM2 foi desenvolvida. As enzimas quiméricas mostraram-se totalmente funcionais e conservaram as propriedades de seus \"domínios\" constituintes. / Introduction and Aims. Phosphatases are promising enzymes for application in the degradation of organophosphates, whereas carbohydrate binding module has significant and demonstrated biotechnological applications. The high-density lipoprotein-associated enzyme paraoxonase 1 (PON1) hydrolyzes lactones, aromatic esters, and neurotoxic organophosphorus (OP) compounds. Our aim is to obtain chimeric proteins containing a phosphatase domain linked to a carbohydrate binding module (CBM), which could be immobilized on a cellulose supports. Results. As a proof of concept, a chimeric protein combining an acid phosphatase (appA) from E.coli and a CBM family 2 (CBM2) from Xanthomonas axonopodis pv. citri was assembled and produced in E.coli as a recombinant soluble protein. Purified appA-CBM2 was fully functional, was bound to microcrystalline cellulose and exhibited phosphatase activity upon p-nitrophenyl phosphate. The binding to microcrystalline cellulose Avicel PH101 exhibited saturation with a binding constant (Kb) of 26 and a maximum binding (Bmax) of 4,45 U/µg. In addition, the binding was higher at pH 2.5 and decreased above pH 6.5, as previously observed for CBM2. Finally, effect of p-nitrophenyl phosphate concentration on appA-CBM2 and native appA activities were identical, exhibiting a Km of 2.8 mM. Taken together, these data show that the conceptual design of a protein combining the properties and biotechnological advantages of phosphatases and cellulose binding domains is possible and functional. Similarly, DNA segments coding for CBM2 and for PON1 from Homo sapiens combined resulting in a segment coding for a chimeric protein (PON1-CBM2). Native PON1 and PON1-CBM2 were produced as recombinant protein in E. coli Arctic. Although purification was not accomplished, these enzymes were characterized. PON1-CBM2 binds to microcrystalline cellulose, exhibiting a saturation behavior described by a Kb of 27 mg. PON1 and PON1- CBM2 have the same Km for paraoxon (1.3 mM), indicating that the phosphatase domain was fully functional. Conclusion. An effective strategy for heterologous expression of the native PON1 and chimeric appA-CBM2 and PON1-CBM2 in E. coli was attained. The chimeric enzymes were fully functional and maintained the properties of their original domains

Introdução e Objetivos. Fosfatases são enzimas promissoras para aplicação na degradação de organofosforados. Por exemplo, a enzima paraoxonase 1 (PON1), associada à lipoproteína de alta densidade (HDL), hidrolisa lactonas, ésteres aromáticos e compostos organofosforados (OP) neurotóxicos. \"Módulos de ligação a carboidrato\" (CBM) têm diversas aplicações biotecnológicas. Nosso objetivo é a obtenção de proteínas quiméricas contendo fosfatases ligadas a um módulo de ligação de celulose, o que possibilitaria a imobilização dessas enzimas em suportes de celulose. Resultados. Como prova de conceito, uma proteína quimérica contendo uma \"fosfatase ácida\" (appA) de E.coli e CBM familia 2 (CBM2) de uma celulase de Xanthomonas axonopodis pv citri foi montada e produzida em E. coli como uma proteína recombinante solúvel. appA-CBM2 purificada demonstrou ser totalmente funcional exibindo atividade de ligação à celulose microcristalina (Avicel PH101) e atividade de fosfatase sobre p-nitrofenil fosfato. A ligação à Avicel evidenciou um comportamento de saturação descrito por uma \"constante de ligação\" (Kb) de 26 mg e um \"máximo de ligação\" (Bmax) de 4,45 U/µg. Além disso, a ligação de appA-CBM2 em Avicel foi maior em pH 2,5 e diminuiu acima de pH 6,5, como observado anteriormente para CBM2. Finalmente, o efeito de concentração de p-nitrofenil fosfato na atividade catalítica de appA-CBM2 e appA foi idêntico, exibindo um Km de 2,8 mM. Portanto, esses dados mostram que o conceito de uma proteína que combina as propriedades da fosfatase e do domínio de ligação à celulose é possível e funcional. De forma similar, os segmentos de DNA que codificam para o CBM2 e para a PON1 de Homo sapiens, foram fusionados resultando em um segmento que codifica para uma proteína quimérica (PON1-CBM2). PON1 nativa e PON1-CBM2 foram produzidas na forma solúvel e ativa em E.coli cepa Arctic. Embora não tenha sido viável sua purificação, estas enzimas foram caracterizadas. PON1-CBM2 liga-se em Avicel PH101 com um comportamento de saturação, descrito por uma constante de ligação (Kb) de 27 mg, valor idêntico àquele observado para appA-CBM2, o que sugere que o domínio CBM2 é igualmente funcional nestas duas enzimas quiméricas. PON1-CBM2 também exibe atividade paraoxonásica com Km similar àquele observado para PON1 nativa (1,3 mM), sugerindo que o \"domínio\" PON1 encontra-se totalmente funcional na enzima quimérica. Conclusão. Uma estratégia para a construção e expressão heteróloga em E. coli de PON1 e das enzimas quiméricas appA-CBM2 e PON1-CBM2 foi desenvolvida. As enzimas quiméricas mostraram-se totalmente funcionais e conservaram as propriedades de seus \"domínios\" constituintes. / Introduction and Aims. Phosphatases are promising enzymes for application in the degradation of organophosphates, whereas carbohydrate binding module has significant and demonstrated biotechnological applications. The high-density lipoprotein-associated enzyme paraoxonase 1 (PON1) hydrolyzes lactones, aromatic esters, and neurotoxic organophosphorus (OP) compounds. Our aim is to obtain chimeric proteins containing a phosphatase domain linked to a carbohydrate binding module (CBM), which could be immobilized on a cellulose supports. Results. As a proof of concept, a chimeric protein combining an acid phosphatase (appA) from E.coli and a CBM family 2 (CBM2) from Xanthomonas axonopodis pv. citri was assembled and produced in E.coli as a recombinant soluble protein. Purified appA-CBM2 was fully functional, was bound to microcrystalline cellulose and exhibited phosphatase activity upon p-nitrophenyl phosphate. The binding to microcrystalline cellulose Avicel PH101 exhibited saturation with a binding constant (Kb) of 26 and a maximum binding (Bmax) of 4,45 U/µg. In addition, the binding was higher at pH 2.5 and decreased above pH 6.5, as previously observed for CBM2. Finally, effect of p-nitrophenyl phosphate concentration on appA-CBM2 and native appA activities were identical, exhibiting a Km of 2.8 mM. Taken together, these data show that the conceptual design of a protein combining the properties and biotechnological advantages of phosphatases and cellulose binding domains is possible and functional. Similarly, DNA segments coding for CBM2 and for PON1 from Homo sapiens combined resulting in a segment coding for a chimeric protein (PON1-CBM2). Native PON1 and PON1-CBM2 were produced as recombinant protein in E. coli Arctic. Although purification was not accomplished, these enzymes were characterized. PON1-CBM2 binds to microcrystalline cellulose, exhibiting a saturation behavior described by a Kb of 27 mg. PON1 and PON1- CBM2 have the same Km for paraoxon (1.3 mM), indicating that the phosphatase domain was fully functional. Conclusion. An effective strategy for heterologous expression of the native PON1 and chimeric appA-CBM2 and PON1-CBM2 in E. coli was attained. The chimeric enzymes were fully functional and maintained the properties of their original domains

This thesis describes strategies for and examples ofcellulose fiber modification.The ability of an engineered biocatalyst, acellulose-binding module fused to theCandida antarcticalipase B, to catalyze ring-openingpolymerization of e-caprolactone in close proximity tocellulose fiber surfaces was explored. The water content in thesystem was found to regulate the polymer molecular weight,whereas the temperature primarily influenced the reaction rate.The hydrophobicity of the cellulose sample increased as aresult of the presence of surface-deposited polyester. A two-step enzymatic method was also investigated. Here,Candida antarctica lipase B catalyzed the acylation ofxyloglucan oligosaccharides.The modified carbohydrates werethen incorporated into longer xyloglucan molecules through theaction of a xyloglucan endotransglycosylase. The modifiedxyloglucan chains were finally deposited on a cellulosesubstrate. The action ofCandida antarcticalipase B was further investigated inthe copolymerization of e-caprolactone and D,L-lactide.Copolymerizations with different e-caprolactone-to-D,L-lactideratios were carried out. Initially, the polymerization wasslowed by the presence of D,L-lactide. During this stage,D,L-lactide was consumed more rapidly than ε-caprolactoneand the incorporation occurred dimer-wise with regard to thelactic acid units. Morphological studies on wood fibers were conducted using asol-gel mineralization method. The replicas produced werestudied, without additional sample preparation, by electronmicroscopy and nitrogen adsorption. Information concerning thestructure and accessibility of the porous fiber wall wasobtained. Studies of never-dried kraft pulp casts revealedmicro-cavities and cellulose fibrils with mean widths of 4.7(±2) and 3.6 (±1) nm, respectively. Finally, cationic catalysis by simple carboxylic acids wasstudied. L-Lactic acid was shown to catalyze the ring-openingpolymerization of ε-caprolactone in bulk at 120 °C.The reaction was initiated with methylß-D-glucopyranoside, sucrose or raffinose, which resultedin carbohydrate-functionalized polyesters. The regioselectivityof the acylation was well in agreement with the correspondinglipase-catalyzed reaction. The polymerization was alsoinitiated with a hexahydroxy-functional compound, whichresulted in a dendrimer-like star polymer. The L-lactic acidwas readily recycled, which made consecutive reactions usingthe same catalyst possible. Keywords:Candida antarcticalipase B, cationic catalysis,cellulose-binding module, dendrimer, enzymatic polymerization,fiber modification, silica-cast replica, sol-gelmineralization, organocatalysis, xyloglucanendotransglycosylase

<p>This thesis describes strategies for and examples ofcellulose fiber modification.The ability of an engineered biocatalyst, acellulose-binding module fused to the<i>Candida antarctica</i>lipase B, to catalyze ring-openingpolymerization of e-caprolactone in close proximity tocellulose fiber surfaces was explored. The water content in thesystem was found to regulate the polymer molecular weight,whereas the temperature primarily influenced the reaction rate.The hydrophobicity of the cellulose sample increased as aresult of the presence of surface-deposited polyester.</p><p>A two-step enzymatic method was also investigated. Here,Candida antarctica lipase B catalyzed the acylation ofxyloglucan oligosaccharides.The modified carbohydrates werethen incorporated into longer xyloglucan molecules through theaction of a xyloglucan endotransglycosylase. The modifiedxyloglucan chains were finally deposited on a cellulosesubstrate.</p><p>The action of<i>Candida antarctica</i>lipase B was further investigated inthe copolymerization of e-caprolactone and D,L-lactide.Copolymerizations with different e-caprolactone-to-D,L-lactideratios were carried out. Initially, the polymerization wasslowed by the presence of D,L-lactide. During this stage,D,L-lactide was consumed more rapidly than ε-caprolactoneand the incorporation occurred dimer-wise with regard to thelactic acid units.</p><p>Morphological studies on wood fibers were conducted using asol-gel mineralization method. The replicas produced werestudied, without additional sample preparation, by electronmicroscopy and nitrogen adsorption. Information concerning thestructure and accessibility of the porous fiber wall wasobtained. Studies of never-dried kraft pulp casts revealedmicro-cavities and cellulose fibrils with mean widths of 4.7(±2) and 3.6 (±1) nm, respectively.</p><p>Finally, cationic catalysis by simple carboxylic acids wasstudied. L-Lactic acid was shown to catalyze the ring-openingpolymerization of ε-caprolactone in bulk at 120 °C.The reaction was initiated with methylß-D-glucopyranoside, sucrose or raffinose, which resultedin carbohydrate-functionalized polyesters. The regioselectivityof the acylation was well in agreement with the correspondinglipase-catalyzed reaction. The polymerization was alsoinitiated with a hexahydroxy-functional compound, whichresulted in a dendrimer-like star polymer. The L-lactic acidwas readily recycled, which made consecutive reactions usingthe same catalyst possible.</p><p><b>Keywords:</b><i>Candida antarctica</i>lipase B, cationic catalysis,cellulose-binding module, dendrimer, enzymatic polymerization,fiber modification, silica-cast replica, sol-gelmineralization, organocatalysis, xyloglucanendotransglycosylase</p>