Abstract Five species belonging to order Chelonia, two families and four genera, namely, Taiwanese soft-shelled turtle, Pelodiscus sinensis japonicus; American soft-shelled turtle, Apalone ferox; alligator snapping turtle, Macroclemys temminck; pitted shelled turtle, Carettochelys insculpta and side-necked turtle, Chelodina siebernrocki were investigated in order to fully understand the structural basis for the multiple lactate dehydrogenase (LDH) isozymes in turtles and soft-shelled turtles. Starch gel electrophoretic patterns of LDH isozymes from muscle, heart, liver, testis and eye were analyzed. Chelonia possess the two fundamental LDH loci-A¡]muscle¡^and B¡]heart¡^as the case of all other vertebrates. The major forms of LDH isozymes in the tissues of Chelonia are homotetrameric LDH-A4 and B4.While some of these Chelonia do not form two heterotetrameric A3B1 and/or A1B3 isozymes. This phenomenon is also observed among some lower vertebrates and fishes of other classes. I have determined the LDH-A and LDH-B cDNA sequences of protein-coding region from these five species. The 3D-structure of tetrameric LDH isozymes from Taiwanese soft-shelled turtle was predicted by homologous modeling and the substitutive residues in subunit contact sites were examined in oder to explain different multiple forms of tetrameric LDH isozymes present in various species. The LDH isozymes are housekeeping genes in most eukaryotic cells and therefore the LDH DNA or protein sequences can be an ideal marker for studying the molecular phylogenetics and evolution among different organisms. However, whether this marker can also be used to investigate the systematic relationship among the closely related species remains to be demonstrated. In this study, the newly determined LDH-A and LDH-B cDNA sequences and their deduced protein sequences from several different turtles and soft-shelled turtles, as well as the previously published LDH sequences, are analyzed by phylogenetic tree reconstruction methods of neighbor-joining, minimum evolution, maximum parsimony and maximum likelihood. These results confirm the traditional classification based on morphology of Chelonia as the two different families belonging to the same order Chelonia. Furthermore, these results clearly classify these Chelonia species into side-necked turtle, rough-shelled hide-necked turtle and smooth-shelled hide-neck turtle. Finally, these results also demonstrate that LDH can indeed be used as a molecular systematic marker for analyzing closely related species.

Thesis (Ph. D.)--University of Hong Kong, 2007. / Title proper from title frame. Also available in printed format.

Thesis (M.S.)--University of New Mexico, 2005. / "July, 2005." Includes bibliographical references (leaves 32-35). FINDMODEL can be accessed with a web-browser.

Thesis (Ph. D.)--University of Aarhus, 2004. / Title from PDF title page (viewed on Jan 3, 2007). Includes articles and manuscripts co-authored with others. Includes bibliographical references. Also available in PostScript format.

Thesis (M.S.)--University of New Mexico, 2005. / "July, 2005." Includes bibliographical references (leaves 50-51). Distplot can be accessed with a web-browser.

Thesis (Ph.D.)--McMaster University, 2005. / Advisor: R. S. Singh. Includes bibliographical references. Also available online.

published_or_final_version / abstract / Biological Sciences / Doctoral / Doctor of Philosophy

One of the most intriguing questions in evolutionary biology is how biochemical and structural complexity arise through small and incremental changes; however answering this question requires an explicit set of candidate residues and an experimental system in which to test them. This dissertation aims to understand how biochemical complexity evolves and assesses the structure-function relationship in the green fluorescent protein (GFP) protein family using an ancestral reconstruction approach. In the second chapter, I studied the evolution of biochemical complexity in Kaede-type red fluorescent proteins (FPs) from Faviina corals. An increase in biochemical complexity is represented by the emergence of red fluorescence because it necessitates the synthesis of a tri-cyclic chromophore from a precursor bi-cyclic chromophore through an additional autocatalytic reaction step. The autocatalytic reaction is fully enabled by as many as twelve historical mutations. Here, I showed that the red fluorescent chromophore evolved from an ancestral green chromophore by perturbing the ancestral protein stability at multiple levels of protein structure. Moreover, only three historical mutations are sufficient to initiate the selection-accessible evolutionary trajectory leading to emergence of red fluorescence. The third chapter investigates six mutations proximate to the chromophore in the Kaede-type FP that could have facilitated autocatalytic synthesis of the red chromophore by enlarging the chromophore-containing cavity and modifying its microenvironment. Two of these six mutations were found to strongly affect the protein’s stability and oligomeric tendency. Additionally, I showed that the dimeric least divergent Kaede-type FP, R1-2, evolved from the tetrameric green ancestor. Taken together the results of these studies indicate that the step-up in biochemical complexity in the Kaede-type FPs was achieved via disruption of the existing stable interactions at tertiary and quaternary protein structure levels. In the fourth chapter, I resurrected the common ancestor of all FPs cloned from the order Leptothecata (class Hydrozoa), which are characterized by the highest known homo-oligomeric diversity. I showed that the ancestor was a green monomeric FP with a large Stokes shift. The ancestral FP together with the extant Leptothecata FPs could server as a model system to study the evolution of function and homo-oligomerization, and the desirable photophysical characteristics would make this ancestral FP a useful bio-marker in bio-medical research. / text

For many species, vision contributes to a number of fitness-related tasks, including mating and the detection of prey and predators. Selection on the visual system should therefore be strong, especially when ecological or genomic changes open new avenues for evolutionary changes. Visual system proteins are thus attractive systems for molecular evolutionary analyses. This thesis presents a collection of evolutionary studies on two gene families, opsins and crystallins. Opsin proteins determine the wavelengths of light detected by the retina, while crystallin proteins contribute to lens transparency and refractory power. My studies focus on teleost fishes, because teleost visual ecology is exceptionally diverse and because gene duplication is common in this group. In Chapter One, I outline the relevance of protein variation to organismal evolution and describe the analytical methods employed throughout this thesis. Chapter Two considers the long-wavelength sensitive (LWS) opsins of the guppy (Poecilia reticulata). The guppy is shown to possess multiple LWS opsins that have accumulated differences at functionally important amino acid sites since duplicating. Chapter Three focuses on the guppy’s main predator, the pike cichlid Crenicichla frenata, which is shown to have a greater capacity for short-wavelength vision than previously believed. However, this cichlid possesses three fewer opsins than closely-related African cichlids, a difference partly due to duplication of a green-sensitive (RH2) opsin in African cichlids. In Chapter Four, this RH2 duplication event is studied in greater depth; variation in selective constraint is documented following gene duplication and between species from different lakes. Some of the analytical methods employed in Chapter Four were newly developed, as detailed in Chapter Five, where a test for functional divergence among clades is evaluated and then improved upon through the presentation of a new null model that better accommodates among-site variation in selection. In Chapter Six, phylogenetic relationships within the βγ lens crystallin superfamily are clarified, and the functionally distinct γN family is shown to have evolved conservatively compared to other crystallin families. The thesis concludes with suggestions for future directions for evolutionary research on opsins and crystallins, and summarizes recent work that has built on these studies.

Nesta tese propomos e estudamos um modelo alternativo para investigar a evolução de quase-espécies moleculares, no qual supomos que a população seja uma combinação aleatória das moléculas constituintes em cada geração. Essa aleatoriedade deve-se a inclusão de um procedimento adicional de amostragem da população além dos procedimentos usuais de mutação e reprodução diferenciada. O modelo, denominado modelo de amostragens, é baseado em um algoritmo que mimetiza procedimentos experimentais que usam técnicas de transferências em série para reproduzir o processo de evolução de microorganismos in vitro. Além do modelo reproduzir a solução exata do estado estacionário do modelo de quase-espécies no regime determinístico, ele permite o estudo da evolução da quase-espécie molecular em todo espaço de parâmetros de controle, incluindo o caso em que a população é finita. A generalização dessa formulação alternativa para uma classe geral de relevos de replicação permite-nos realizar um estudo bastante completo do fenômeno do limiar de erro, levando-nos a uma análise crítica sobre a generalidade desse fenômeno / In this thesis we propose and study an alternative model to investigate the evolution of a molecular quasispecies, in which we assume that the population is a random combination of the constituent molecules in each generation. This randomness is due to the inclusion of an additional sampling procedure of the population, besides the usual procedures of mutation and differential reprodution. This model, termed sampling model, is based on an algorithm that mimics experimental procedures using serial transfer techniques to study the microbial evolutionary process in vitro. Besides yielding the exact steady-state solution of the quasispecies model in the deterministic limit, the sampling model allows the study of the molecular evolution in the full space of the control parameter, including the case where of population is finite. The generalization of this alternative formulation to a general class of fitness landscapes, allows us to investigate thoroughly the error threshold phenomenon, leading us to discuss critically the generality of this phenomenon in molecular evolution