The essence of metabolic engineering is the modification of microbes for the overproduction of useful compounds. These cellular factories are increasingly recognized as an environmentally-friendly and cost-effective way to convert inexpensive and renewable feedstocks into products, compared to traditional chemical synthesis from petrochemicals. The products span the spectrum of specialty, fine or bulk chemicals, with uses such as pharmaceuticals, nutraceuticals, flavors and fragrances, agrochemicals, biofuels and building blocks for other compounds. However, the process of metabolic engineering can be long and expensive, primarily due to technological hurdles, our incomplete understanding of biology, as well as redundancies and limitations built into the natural program of living cells. Combinatorial or directed evolution approaches can enable us to make progress even without a full understanding of the cell, and can also lead to the discovery of new knowledge. This thesis is focused on addressing the technological bottlenecks in the directed evolution cycle, specifically de novo DNA assembly to generate strain libraries and small molecule product screens and selections.

Thesis (M.S.)--Bowling Green State University, 2009. / Document formatted into pages; contains viii, 38 p. : ill. Includes bibliographical references.

Zymomonas mobilis is a non-pathogenic, facultatively-anaerobic Gram-negative bacterium, which has historically been used for the fermentation of alcoholic beverages in many tropical/sub-tropical countries. Due to its excellent ethanol-producing capabilities, significant effort has been undertaken over recent years to utilize it for industrial ‘bioethanol’ production. Its physiological and metabolic properties indicate that it may also be an excellent organism for the bio-production of many different types of organic molecules. Consequently, the aim of my thesis was to develop new molecular methodologies that would enable Z. mobilis to be ‘engineered’ for use in future ‘bioproduction’ endeavours. In the first part of my study, I analyzed the native (cryptic) plasmids present within a variety of Z. mobilis strains, including two poorly-studied Z. mobilis strains: NCIMB 11163 and NCIMB 8227. Several plasmid libraries containing restriction-digested fragments of Z. mobilis cryptic plasmid DNA were prepared, and their inserts were sequenced. This enabled the complete DNA sequences of three small (non-integrating, double-stranded DNA) cryptic plasmids to be determined: pZMO1A and pZMO7 from NCIMB 11163, and pZMO1B from NCIMB 8227. Their DNA sequences were analyzed using bioinformatic approaches, to identify open reading frames, and regions of DNA that were putatively involved in transcription or DNA replication. In the second part of this thesis, the minimally-replicating region from plasmid pZMO7 was used to construct a series of Escherichia coli-Z. mobilis shuttle vectors. These vectors were found to be stable within several Z. mobilis strains for over 60 generations without antibiotic selective pressure. A reliable and reproducible method based on quantitative real time PCR (Q-RT-PCR) was developed to accurately determine the copy number of cryptic plasmids and shuttle-vectors present in Z. mobilis cultures. The pZMO7-based shuttle vectors exhibited good compatibility with cryptic plasmids as well as the widely-used pZM2-based shuttle vectors. Genes encoding glutathione S-transferase (GST) as well as green and red fluorescent protein (GFP and RFP) reporters were cloned into various shuttle vector constructs; placing them under the control of endogenous (Ppdc) or exogenous (Plac and Ptac) promoters. Promoter strength was evaluated by quantifying the reporter gene expression. The plasmid-based expression of GFP and RFP was visualized within planktonic and biofilm cultures using confocal laser scanning microscopy (CLSM). Shuttle vector-based GST pull-down experiments were used to study intracellular protein-protein binding interactions. In the third part of my thesis, I explored the potential use of Z. mobilis for the bioproduction of isoprenoid (terpenoid) compounds. Five predicted sesquiterpene synthases (terpene cyclases) of unknown function from the dimorphic fungus Penicillium marneffei, and several terpene cyclases from several other bacteria, fungi and plants were initially functionally-analyzed in E. coli. Several cyclase genes were cloned into E. coli-Z. mobilis shuttle vectors for expression trials within Z. mobilis cells. In summary, this thesis describes the development of a variety of novel methodologies and genetic ‘tools’ that may be used to express heterologous genes within Z. mobilis cells. These will be invaluable for future studies concerned with exploring the biology and industrial applications of this ‘microbial cell factory’. / published_or_final_version / Chemistry / Doctoral / Doctor of Philosophy

Five transposon Tn5 mutants of bean rhizobia (Rhizobium leguminosarum b.v. phaseoli) and the wild type strain were used in ecological studies to evaluate the efficacy of transposon Tn5 as a phenotypic marker in rhizobia for ecological studies in two Sonoran desert soils. All mutants possessed chromosomal insertions of the transposable element. Survival of each mutant strain was compared to that of the wild type strain under non stress, moisture stress and temperature stress conditions in Pima silty clay loam and Brazil to sandy loam. The genetic stability of Tn5 in terms of transposition of the element within the chromosome and the Tn5 coded antibiotic resistant phenotype was determined in cells recovered throughout the survival period. Under non stress conditions, the viable Tn5 mutant population decreased in size. Two mutants showed significantly (p < 0.01) lower populations than the wild type at the end of 30 days in the silty clay loam. In the sandy loam, four of the five mutant populations were significantly lower than the wild type. Tn5 was genetically stable in both soils. Under moisture stress conditions, the decline of the Tn5 mutant and wild type populations corresponded to a decline in soil moisture content. The finer textured soil afforded more protection to the cells than the coarse textured soil. There were no indications of Tn5 instability under moisture stress. In both soils under temperature stress, sizes of all populations declined rapidly and after 12 days, the mutant cells when screened using the Tn5 coded markers were significantly less in numbers than the wild type indicating a loss of Tn5 coded antibiotic resistance phenotype. There were no significant differences in numbers between wild type and mutant cells when screened using only the intrinsic markers. DNA:DNA hybridizations confirmed that the lack of Tn5 coded antibiotic resistance phenotype was probably not due to a deletion or transposition of the element. Under non stress conditions Tn5 is a useful ecological marker, but each Tn5 mutant has to be evaluated independently under specific environmental conditions to determine the efficacy of Tn5 as an ecological marker.

Environmental use of genetically-engineered microorganisms (GEMs) has raised concerns over potential ecological impact. Development of microcosm systems useful in preliminary testing for risk assessment will provide useful information for predicting potential structural, functional, and genetic effects of GEM release. This study was executed to develop techniques that may be useful in risk assessment and microbial ecology, to ascertain which parameters are useful in determining risk and to predict risk from releasing an engineered strain of Erwinia carotovora. A terrestrial microcosm system for use in GEM risk assessment studies was developed for use in assessing alterations of microbial structure and function that may be caused by introducing the engineered strain of E. carotovora. This strain is being developed for use as a biological control agent for plant soft rot. Parameters that were monitored included survival and intraspecific competition of E. carotovora, structural effects upon both total bacterial populations and numbers of selected bacterial genera, effects upon activities of dehydrogenase and alkaline phosphatase, effects upon soil nutrients, and potential for gene transfer into or out of the engineered strain. No significant difference was found in survival of the engineered strain as compared to its wildtype parent. Both strains survived for over two months in microcosms. The effects of both strains upon populations of total bacteria and selected bacterial genera were determined; while some effects upon community structure were observed, they were not significant. The engineered strain was not found to be a superior competitor compared to its parent; three different doses of engineered and wildtype strains were used. ln addition, neither strain affected activities of dehydrogenase or alkaline phosphatase in soil. Likewise, no effects were observed upon the nutrients monitored. However, transfer of the kanamycin resistance gene that had been inserted into the engineered E. carotovora strain may have occurred. Five species of indigenous bacteria displayed kanamycin resistance 15 days after being exposed to the engineered Erwinia. DNA from these strains was isolated, purified, and hybridization experiments executed to determine if any homology existed between these DNAs and the kanamycin resistance gene that had been inserted into E. carotovora. Using biotin-Iabeled probes and Iow-stringency washing conditions, homology was observed. However, before gene transfer can be proven, additional studies, including amplification and sequencing, may be required. Although a simple microcosm design was employed, it yielded sufficient information to conclude that release of the engineered Erwinia carotovora will not affect any of the microbial measures of integrity that were studied in a manner different from that of the wildtype. Effects upon plant material and other higher taxa will be the focus of future studies. / Ph. D.

The ability of a genetically-engineered <i>Erwirzia carotovora</i> subsp. <i>carotovora</I> (Ecc) strain to transfer recombinant chromosomal DNA or plasmids to wildtype Ecc or <i>Pseudomonas fluorescens</i> was tested on filters, within soil microcosms, and <i>in planta</i>. Ecc was engineered by chromosomal insertion of a disarmed <i>endo</i>-pectate lyase gene marked with a 1.4kb DNA fragment conferring kanamycin resistance. Plasmids RPI and pBR322 were introduced separately into engineered Ecc clones. These strains served as donors in genetic transfer experiments. No transfer of the inserted kan marker or of pBR322 was observed under any experimental condition. In filter matings, RPI was transferred to wildtype Ecc at a frequency of 3.6 X 10⁻² transconjugants per donor (TPD) and to P. <i>fluorescens</i> at a frequency of 2.4 X 10⁻⁵ TPD. In matings conducted in potato tubers inoculated using sewing needles, the respective frequencies were 4.0 X 10⁻³ and 2.0 X 10⁻³, while matings on potato slices yielded frequencies of 4.7 X 10⁻² and 2.3 X 10⁻². In soil microcosms, the maximum transfer frequencies observed were 2.3 X 10³ and 8.4 X 10⁻⁵ TPD. / Master of Science

O fungo basidiomiceto Crinipellis perniciosa é o agente causal da vassoura-de-bruxa do cacaueiro (Theobroma cacao), podendo causar sérios perdas na produção. Este fungo é capaz de colonizar, além do cacau, várias outras plantas hospedeiras, onde pode se adaptar às novas condições, contribuindo, dessa forma, para o aumento da variabilidade genética desse microrganismo. Assim sendo, o objetivo desse trabalho foi isolar e identificar C. perniciosa de tecidos sadios de cacaueiro e comparar a sua variabilidade genética e fisiológica com isolados patogênicos por meio de análises moleculares e fisiológicas. Após a desinfecção e retirada da casca dos ramos, colônias morfologicamente similares a C. perniciosa foram obtidas de tecidos sadios de cacaueiro, os quais foram considerados como endófitos. A identificação preliminar desses isolados foi baseada nas observações morfológicas, tais como coloração da colônia e presença de grampos de conexão. Posteriormente, os isolados de C. perniciosa foram caracterizados geneticamente por RAPD, e sequenciamento de regiões do rDNA (18S+5.8S+28S), resistência a fungicidas, produção de exoenzimas e patogenicidade ao cacaueiro. A análise por marcadores RAPD separou os 37 isolados em 8 grupos (G1 a G8), sendo o grupo G1 dividido em 2 subgrupos (G1-1 e G1-2). Por esta análise foi possível observar que linhagens com 100 % de similaridade foram isoladas de locais diferentes, enquanto que isolados obtidos de uma mesma planta podem ser geneticamente diferentes. A análise da seqüência do rDNA dos isolados de C. perniciosa mostrou a formação de 6 grupos distintos (R1 a R6). Isolados obtidos de uma mesma planta não se agruparam, reforçando a hipótese de que isolados geneticamente diferentes podem ocupar a mesma planta hospedeira. Quanto resistência a fungicidas (tebuconazole, mancozeb, benomil e óxido cuproso), foi observado que tebuconazole foi o mais eficiente na inibição do crescimento miceliar, enquanto benomil apresentar menor taxa de inibição. Isolados endofíticos e patogênicos apresentaram comportamento similar em relação aos fungicidas. Resultado semelhante foi observado quanto a produção de exoenzimas (amilase, lipase, pectinase, exo e endoglicanase). Foi observada a produção de todas as enzimas avaliadas, sendo possível detectar variações entre os isolados. Entretanto, esta variação não foi correlacionada ao hábito endofítico ou patogênico. A patogenicidade de 8 isolados de C. perniciosa foi avaliada em mudas de cacau Catongo. A porcentagem de plantas infectadas com sintomas variou de 55% a 100%. Isolados endofíticos apresentaram baixa virulência sobre o cacaueiro, tendo o isolado endofítico 31 induziu sintomas em 60% das plantas inoculadas, resultado este provavelmente devido à alta pressão de inóculo. Este trabalho mostra pela primeira vez C. perniciosa como endófito em tecidos não meristemático do cacaueiro. / The basidiomycete fungus Crinipellis perniciosa (Stahel) Singer is the causal agent of Witches' Broom Disease of Cacao (Theobroma cacao L.) which is the main factor limiting cocoa production in the Americas. Pod losses of up to 90% are experienced in affected areas as evidenced by the 50% drop in production in Bahia province, Brazil following the arrival of the C. perniciosa in the area in 1989. The disease has proven particularly difficult to control and many farmers in affected areas have given up cacao cultivation. Any useful control strategy for Witches' Broom disease must be effective against in range strains of the pathogen. It is already known that pathogenic variation exists among isolates of Crinipellis perniciosa obtained from different areas and host plants. However, any study was developed about the variability between endophytic and pathogenic population. In order to evaluate the genetic variability of 35 isolates of endophytic and pathogenic populations of Crinipellis perniciosa, the RAPD technique, ITS sequencing and fungicide susceptibility were performed. Genetic variability between 35 isolates of C. perniciosa was analysed by the random amplified polymorphic DNA (RAPD) technique, which indicated that isolates from other host plants were more diverse than isolates obtained from cacao plants. Among cacao isolates was observed at least two groups, one formed mainly by endophytic isolates and other by pathogenic ones. Analysis by ITS sequence grouped isolates independently of endophytic or pathogenic status. Fungicide susceptibility showed that cupric oxide inhibits statistically more endophytic isolates than pathogenic ones, showing that could have physiological differences between these populations. The present study highlighted the possible genetic and physiological differences between endophytic and pathogenic population of C. perniciosa.

O fungo basidiomiceto Crinipellis perniciosa é o agente causal da vassoura-de-bruxa do cacaueiro (Theobroma cacao), podendo causar sérios perdas na produção. Este fungo é capaz de colonizar, além do cacau, várias outras plantas hospedeiras, onde pode se adaptar às novas condições, contribuindo, dessa forma, para o aumento da variabilidade genética desse microrganismo. Assim sendo, o objetivo desse trabalho foi isolar e identificar C. perniciosa de tecidos sadios de cacaueiro e comparar a sua variabilidade genética e fisiológica com isolados patogênicos por meio de análises moleculares e fisiológicas. Após a desinfecção e retirada da casca dos ramos, colônias morfologicamente similares a C. perniciosa foram obtidas de tecidos sadios de cacaueiro, os quais foram considerados como endófitos. A identificação preliminar desses isolados foi baseada nas observações morfológicas, tais como coloração da colônia e presença de grampos de conexão. Posteriormente, os isolados de C. perniciosa foram caracterizados geneticamente por RAPD, e sequenciamento de regiões do rDNA (18S+5.8S+28S), resistência a fungicidas, produção de exoenzimas e patogenicidade ao cacaueiro. A análise por marcadores RAPD separou os 37 isolados em 8 grupos (G1 a G8), sendo o grupo G1 dividido em 2 subgrupos (G1-1 e G1-2). Por esta análise foi possível observar que linhagens com 100 % de similaridade foram isoladas de locais diferentes, enquanto que isolados obtidos de uma mesma planta podem ser geneticamente diferentes. A análise da seqüência do rDNA dos isolados de C. perniciosa mostrou a formação de 6 grupos distintos (R1 a R6). Isolados obtidos de uma mesma planta não se agruparam, reforçando a hipótese de que isolados geneticamente diferentes podem ocupar a mesma planta hospedeira. Quanto resistência a fungicidas (tebuconazole, mancozeb, benomil e óxido cuproso), foi observado que tebuconazole foi o mais eficiente na inibição do crescimento miceliar, enquanto benomil apresentar menor taxa de inibição. Isolados endofíticos e patogênicos apresentaram comportamento similar em relação aos fungicidas. Resultado semelhante foi observado quanto a produção de exoenzimas (amilase, lipase, pectinase, exo e endoglicanase). Foi observada a produção de todas as enzimas avaliadas, sendo possível detectar variações entre os isolados. Entretanto, esta variação não foi correlacionada ao hábito endofítico ou patogênico. A patogenicidade de 8 isolados de C. perniciosa foi avaliada em mudas de cacau Catongo. A porcentagem de plantas infectadas com sintomas variou de 55% a 100%. Isolados endofíticos apresentaram baixa virulência sobre o cacaueiro, tendo o isolado endofítico 31 induziu sintomas em 60% das plantas inoculadas, resultado este provavelmente devido à alta pressão de inóculo. Este trabalho mostra pela primeira vez C. perniciosa como endófito em tecidos não meristemático do cacaueiro. / The basidiomycete fungus Crinipellis perniciosa (Stahel) Singer is the causal agent of Witches' Broom Disease of Cacao (Theobroma cacao L.) which is the main factor limiting cocoa production in the Americas. Pod losses of up to 90% are experienced in affected areas as evidenced by the 50% drop in production in Bahia province, Brazil following the arrival of the C. perniciosa in the area in 1989. The disease has proven particularly difficult to control and many farmers in affected areas have given up cacao cultivation. Any useful control strategy for Witches' Broom disease must be effective against in range strains of the pathogen. It is already known that pathogenic variation exists among isolates of Crinipellis perniciosa obtained from different areas and host plants. However, any study was developed about the variability between endophytic and pathogenic population. In order to evaluate the genetic variability of 35 isolates of endophytic and pathogenic populations of Crinipellis perniciosa, the RAPD technique, ITS sequencing and fungicide susceptibility were performed. Genetic variability between 35 isolates of C. perniciosa was analysed by the random amplified polymorphic DNA (RAPD) technique, which indicated that isolates from other host plants were more diverse than isolates obtained from cacao plants. Among cacao isolates was observed at least two groups, one formed mainly by endophytic isolates and other by pathogenic ones. Analysis by ITS sequence grouped isolates independently of endophytic or pathogenic status. Fungicide susceptibility showed that cupric oxide inhibits statistically more endophytic isolates than pathogenic ones, showing that could have physiological differences between these populations. The present study highlighted the possible genetic and physiological differences between endophytic and pathogenic population of C. perniciosa.