Inexpensive petroleum is the cornerstone of the modern global economy despite its huge environmental costs and its nature as a non-renewable resource. While ninety percent of petroleum is ultimately used as fuel and can in principle be replaced by sources of renewable electricity, ten percent is used as a feedstock to produce societally important chemicals that cannot currently be made at a reasonable cost through alternative processes. In this dissertation, I will discuss my efforts, together with several colleagues, to apply synthetic biology approaches to the challenge of producing renewable petrochemical replacements. In Chapter 2, I discuss our efforts to engineer E. coli to produce fatty acids with a wide range of chain lengths at high yield, thereby providing an alternative platform for the production of diverse petrochemicals. In Chapter 3, I describe a novel method of DNA assembly that we developed to facilitate synthetic biology efforts such as those in Chapter 2. This method is capable of simultaneously assembling multiple DNA pieces with substantial sequence homology, a common challenge in synthetic biology. In Chapter 4, I discuss the development of a "bionic leaf": a hybrid microbial-inorganic catalyst that marries the advantages of photovoltaic-based light capture and microbial carbon fixation to achieve solar biomass yields greater than those observed in terrestrial plants. This technology offers a potentially low-cost alternative to photosynthesis as a source of biomass and derived chemicals and fuels. The work described in this dissertation demonstrates the capacity of synthetic biology to address the problem of renewable chemical production, and offers proof of principle demonstrations that both the scope and efficiency of biological chemical production may be improved.

In this work, we seek to expand synthetic in vitro biological systems by using water-in-oil emulsions to provide an environment conducive to directed evolution. We approach this primarily by utilizing a model transcription system, the T7 RNA polymerase and promoter, which is orthogonal to both bacterial and eukaryotic transcription systems and is highly functional in vitro. First, we develop a method to identify functional promoter sequences completely in vitro. This method is tested using the T7 RNA polymerase-promoter model system. We then configure the T7 transcription system as an ‘autogene’ and investigate how this positive feedback circuit (whereby a T7 promoter expresses a T7 RNA polymerase gene) functions across various in vitro platforms, including while compartmentalized. The T7 autogene can be envisioned as a self-replicating system when compartmentalized, and its use for directed evolution is examined. Finally, we look towards future uses for these in vitro systems. One interesting application is to expand the utilization of unnatural base pairs within the context of a synthetic system. We investigate the ability of T7 RNA polymerase to recognize and utilize unnatural base pairs within the T7 promoter, complementing existing work on the utilization of unnatural base pairs for in vitro replication and transcription with an investigation of more complex protein-dependent regulatory function. We envision this work as a foundation for future in vitro synthetic biology efforts. / text

<p>Diverse mechanisms have been proposed to explain natural pattern formation processes, such as slime mold aggregation, feather branching, and tissue stratification. Regardless of the specific molecular interactions, the vast majority of these mechanisms invoke morphogen gradients, which are either predefined or generated as part of the patterning processes. However, using E. coli programmed by a simple synthetic gene circuit, I demonstrate here the generation of robust, self-organized ring patterns of gene expression in the absence of an apparent morphogen gradient. Interestingly, modeling and experimental tests show that the temporal dynamics of the global morphogen concentration serve as a timing mechanism to trigger formation and maintenance of these ring patterns, which are readily tunable by experimentally controllable environmental factors. This mechanism represents a novel mode of pattern formation that has implications for understanding natural developmental processes. In addition, the system can be coupled with inkjet printing technology and metabolic engineering approaches to develop future complex patterned biomaterials.</p> / Dissertation

<p>Various types of genetic constructs are widely used as diagnostic, prophylactic, and therapeutic tools for human diseases. They are also the workhorse in biotech and pharmaceutical industry for production of therapeutic antibodies and proteins. Since the majority of the genetic constructs encode protein products, it is therefore of tremendous value to human health and the society that we could find a way to fine-tune and optimize genetic constructs and hence protein expression for achieving maximal potency or long-lasting effects in therapeutics or for obtaining highest yields in pharmaceutical protein production. However, for protein-coding genes to be expressed in a heterologous host, the coding sequences need to be optimized by using synonymous codons to achieve reasonable levels of expression, if at all. Since codon optimization is done in a protein-by-protein basis with respect to specific host organisms, tissue/cell types, even health conditions, and there is no set of standard rules to follow, this process is still very unpredictable and time-consuming.</p><p>This thesis presents the development of a feasible platform for solving the problem of optimizing regular and long DNA constructs for academic or industrial purposes through the development of a novel cloning method for complex gene libraries, and based on the library expression system constructed in such manner, a platform for high-throughput screening of codon-optimized and error-corrected proteins, and a novel protocol for screening long gene constructs which could be extremely difficult to achieve by using regular screening methods. This multi-step platform has the potential for studying the natural systems: how codon bias correlates to protein expression efficiency, for generating improved pharmaceutical proteins and enhanced DNA vaccines and for constructing improved genome libraries.</p> / Dissertation

Two major roadblocks in synthetic biology are the difficulties associated with the physical assembly of gene regulatory networks (GRNs) and the lack of characterized biological parts. In this work we aimed to address both of these issues. We developed a novel method for the assembly of GRNs called Brick- Mason assembly. We have shown that the method can assemble a 6 part network in a single day and provides significant advancements over traditional cloning methods. We used BrickMason to assemble GRNs that would allow us to compare natural yeast mechanisms of repression to the steric hindrance based mechanisms that are commonly used in synthetic GRNs in yeast. Our results show that the two mechansisms of repression are not equivalent. This finding opens possibilities for using a new class of repressor in a synthetic context in yeast.

Living cells are exquisite systems. They are strongly regulated to perform in highly specific ways, but are at the same time wonderfully robust. This combination arises from the sophistication of their construction and operation: their internal variables are carefully controlled by complex networks of dynamic biochemical interactions, crafted and refined by billions of years of evolution. Using mod- ern DNA engineering technology, scientists have begun to circumvent the long process of evolution by employing a rational design-based approach to construct novel gene networks inside living cells. Currently, these synthetic networks are relatively simple when compared to their natural counter- parts, but future prospects are promising, and synthetic biologists would one day like to be able to control cells using genetic circuits much in the way that electronic devices are controlled using electrical circuits. The importance of precise dynamical behaviour in living organisms suggests that this endeavour would benefit greatly from the insights of control theory. However, the nature of bio- chemical networks can make the implementation of even basic control structures challenging. This thesis focusses specifically on the concept of integral control in this context. Integral control is a fun- damental strategy in control theory that is central to regulation, sensory adaptation, and long-term robustness. Consequently, its implementation in a synthetic gene network is an attractive prospect. Here, the general challenges and important design considerations associated with engineering an in-cell synthetic integral controller are laid out. Specific implementations using transcriptional regu- lation are studied analytically and then in silico using models constructed with commonly available parts from the bacterium Escherichia coli. Finally, using a controller based on post-translational signalling, an on-paper design is proposed for an integral-controlled biosynthesis network intended to allow a population of engineered Saccharomyces cerevisiae cells to actively regulate the extracellular concentration of a small molecule.

Living cells are exquisite systems. They are strongly regulated to perform in highly specific ways, but are at the same time wonderfully robust. This combination arises from the sophistication of their construction and operation: their internal variables are carefully controlled by complex networks of dynamic biochemical interactions, crafted and refined by billions of years of evolution. Using mod- ern DNA engineering technology, scientists have begun to circumvent the long process of evolution by employing a rational design-based approach to construct novel gene networks inside living cells. Currently, these synthetic networks are relatively simple when compared to their natural counter- parts, but future prospects are promising, and synthetic biologists would one day like to be able to control cells using genetic circuits much in the way that electronic devices are controlled using electrical circuits. The importance of precise dynamical behaviour in living organisms suggests that this endeavour would benefit greatly from the insights of control theory. However, the nature of bio- chemical networks can make the implementation of even basic control structures challenging. This thesis focusses specifically on the concept of integral control in this context. Integral control is a fun- damental strategy in control theory that is central to regulation, sensory adaptation, and long-term robustness. Consequently, its implementation in a synthetic gene network is an attractive prospect. Here, the general challenges and important design considerations associated with engineering an in-cell synthetic integral controller are laid out. Specific implementations using transcriptional regu- lation are studied analytically and then in silico using models constructed with commonly available parts from the bacterium Escherichia coli. Finally, using a controller based on post-translational signalling, an on-paper design is proposed for an integral-controlled biosynthesis network intended to allow a population of engineered Saccharomyces cerevisiae cells to actively regulate the extracellular concentration of a small molecule.

Synthetic biology is a fairly recent field that aims to engineer novel functions in biological systems. In a broad sense synthetic biology encompasses the development of tools that makes the engineering of biology easier. In this thesis I develop a collection of standard DNA parts (Biobricks) that consists of a tool to build custom eukaryotic plasmids. This is not just intended for biology researchers in the field of synthetic biology, but also for more general use. Besides the development of molecular biology tools that facilitate the engineering of biology, synthetic biology researchers have implemented devices that are electronics-like in behavior and have demonstrated the potential of the field for the production of biofuels, pharmaceutics and biosensors. Here I present a sensor of iron regulatory protein activity, based on Biobricks. To demonstrate its use I apply it to the study of a novel reconstituted two cell-type co-culture (BNL CL.2 and RAW 264.7), surrogate for hepatocyte-macrophage communication / La biología sintética es un campo recientemente desarrollado con el objectivo de implementar nuevas funciones en sistemas biológicos. De forma global, la biología sintética incluye el desarrollo de herramientas para facilitar la ingeniería de sistemas biológicos. En diversas publicaciones, investigadores en el campo de la biología sintética han implementado dispositivos que funcionan de forma similar a circuitos electrónicos y han demonstrado el potencial del campo para la producción de biocarburantes, farmaceuticos y biosensores. Para la presente tesis he creado una colección de plasmidos estandarizados (Biobricks) que pueden ser de interés para biólogos fuera del campo da la biología sintética. Además, utilizando estos Biobricks, he creado un sensor de la actividad de las proteínas reguladas por el hierro. Para demonstrar su aplicación, he utilizado el sensor para estudiar un nuevo sistema de co-cultura de dos tipos celulares (BNL CL.2 y RAW 264.7), substituto para la comunicación entre hepatocitos y macrófagos

p. 1-65 / Submitted by Antonio Geraldo Couto Barreto (ppgms@ufba.br) on 2013-10-08T13:13:03Z No. of bitstreams: 1 TESE_Luciana_VERSÃO FINAL.pdf: 730192 bytes, checksum: 547f5cb7e2a72752f841739831892186 (MD5) / Approved for entry into archive by Patricia Barroso(pbarroso@ufba.br) on 2013-10-08T16:57:30Z (GMT) No. of bitstreams: 1 TESE_Luciana_VERSÃO FINAL.pdf: 730192 bytes, checksum: 547f5cb7e2a72752f841739831892186 (MD5) / Made available in DSpace on 2013-10-08T16:57:30Z (GMT). No. of bitstreams: 1 TESE_Luciana_VERSÃO FINAL.pdf: 730192 bytes, checksum: 547f5cb7e2a72752f841739831892186 (MD5) Previous issue date: 2013 / A Leishmaniose visceral (LV) é uma doença crônica, endêmica em 62 países e representa um sério problema de saúde pública no Brasil. Os testes sorodiagnósticos convencionais empregam antígenos inteiros ou extratos solúveis que limitam a padronização do antígeno, e podem gerar reações cruzadas com outras doenças. Um método alternativo é o uso de peptídeos a partir de epitopos de célula B identificados através de ferramentas de bioinformática. Objetivou-se identificar epitopos lineares e conformacionais de célula B das proteínas de Leishmania infantum cisteína peptidase calpaina-like, redutase thiol dependente 1 (TDR1) e HSP70, bem como identificar sua estrutura secundária através de metodologia in silico; em seguida, buscou-se selecionar os epitopos lineares comuns aos diferentes métodos de predição para verificar a composição dos resíduos de aminoácidos dos mesmos. Metodologia: As ferramentas de bioinformática IEDB, BepiPred e BcePred foram usadas para predição de epitopos lineares de célula B e o programa CBtope para predição de epitopos conformacionais. A estrutura secundária das proteínas foi predita pelo servidor PHD. Resultados: As análises de predição produziram um total de 148 epitopos lineares e 164 epitopos conformacionais a partir das três proteínas, a maioria desses epitopos está localizada na mesma região. A estrutura secundária das proteínas é composta por -hélice, fita estendida e randômica. Nas proteínas TDR1 e HSP70, os epitopos preditos estão localizados principalmente em regiões de -hélice e randômica. Conclusões: Epitopos lineares e conformacionais de célula B de proteínas de L. infantum foram identificados in silico e poderão contribuir como novos antígenos com potencial aplicação no diagnóstico e controle da leishmaniose visceral. Sugere-se que vários métodos de predição de epitopos lineares sejam combinados a fim de se obter resultados mais confiáveis. / Salvador

p. 1-65 / Submitted by Antonio Geraldo Couto Barreto (ppgms@ufba.br) on 2013-10-04T11:30:48Z No. of bitstreams: 1 TESE_Luciana_VERSÃO FINAL.pdf: 730192 bytes, checksum: 547f5cb7e2a72752f841739831892186 (MD5) / Approved for entry into archive by Patricia Barroso (pbarroso@ufba.br) on 2013-10-30T19:39:53Z (GMT) No. of bitstreams: 1 TESE_Luciana_VERSÃO FINAL.pdf: 730192 bytes, checksum: 547f5cb7e2a72752f841739831892186 (MD5) / Made available in DSpace on 2013-10-30T19:39:53Z (GMT). No. of bitstreams: 1 TESE_Luciana_VERSÃO FINAL.pdf: 730192 bytes, checksum: 547f5cb7e2a72752f841739831892186 (MD5) Previous issue date: 2013 / A Leishmaniose visceral (LV) é uma doença crônica, endêmica em 62 países e representa um sério problema de saúde pública no Brasil. Os testes sorodiagnósticos convencionais empregam antígenos inteiros ou extratos solúveis que limitam a padronização do antígeno, e podem gerar reações cruzadas com outras doenças. Um método alternativo é o uso de peptídeos a partir de epitopos de célula B identificados através de ferramentas de bioinformática. Objetivou-se identificar epitopos lineares e conformacionais de célula B das proteínas de Leishmania infantum cisteína peptidase calpaina-like, redutase thiol dependente 1 (TDR1) e HSP70, bem como identificar sua estrutura secundária através de metodologia in silico; em seguida, buscou-se selecionar os epitopos lineares comuns aos diferentes métodos de predição para verificar a composição dos resíduos de aminoácidos dos mesmos. Metodologia: As ferramentas de bioinformática IEDB, BepiPred e BcePred foram usadas para predição de epitopos lineares de célula B e o programa CBtope para predição de epitopos conformacionais. A estrutura secundária das proteínas foi predita pelo servidor PHD. Resultados: As análises de predição produziram um total de 148 epitopos lineares e 164 epitopos conformacionais a partir das três proteínas, a maioria desses epitopos está localizada na mesma região. A estrutura secundária das proteínas é composta por -hélice, fita estendida e randômica. Nas proteínas TDR1 e HSP70, os epitopos preditos estão localizados principalmente em regiões de -hélice e randômica. Conclusões: Epitopos lineares e conformacionais de célula B de proteínas de L. infantum foram identificados in silico e poderão contribuir como novos antígenos com potencial aplicação no diagnóstico e controle da leishmaniose visceral. Sugere-se que vários métodos de predição de epitopos lineares sejam combinados a fim de se obter resultados mais confiáveis. / Salvador