UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Molecular phylogenetics of Trichonympha cf. collaris and a putative pyrsonymphid: the relevance to the… Dacks, Joel Bryan 1998

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


831-ubc_1998-0215.pdf [ 9.67MB ]
JSON: 831-1.0088401.json
JSON-LD: 831-1.0088401-ld.json
RDF/XML (Pretty): 831-1.0088401-rdf.xml
RDF/JSON: 831-1.0088401-rdf.json
Turtle: 831-1.0088401-turtle.txt
N-Triples: 831-1.0088401-rdf-ntriples.txt
Original Record: 831-1.0088401-source.json
Full Text

Full Text

MOLECULAR PHYLOGENETICS OF TRICHONYMPHA CF. COLLARIS A N D A PUTATIVE PYRSONYMPHID: T H E R E L E V A N C E TO THE ORIGIN OF SEX by JOEL BRYAN DACKS B.Sc. The University of Alberta, 1995 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE O F MASTER'S OF SCIENCE in THE FACULTY OF GRADUATE STUDIES (Department of Zoology) We accept this thesis as conforming to the required standard  THE UNIVERSITY OF BRITISH COLUMBIA April 1998 © Joel Bryan Dacks,  1998  In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission.  Department of  ~2—oc)^Oa^  The University of British Columbia Vancouver, Canada Date {X^ZY  DE-6 (2/88)  Z- V.  /^P  Abstract Why sex evolved is one of the central questions in evolutionary genetics. To address this question I have undertaken a molecular phylogenetic study of two candidate lineages to determine the first sexual line. In my thesis the hypermastigotes are confirmed as closely related to the trichomonads in the phylum Parabasalia and found to be more deeply divergent than a putative pyrsonymphid. This means that the Parabasalia are the first sexual lineage. From this I go on to infer that the ancestral sexual cycle included facultative sex. The relevance of these inferences is then examined with respect to the current theories on the origin of sex.  ii  Table of contents Abstract  •  ii  List of Tables  iv-  List of Figures  :  vii  Acknowledgements CHAPTER I  v  Introduction  1  General question  1  Theories on the origin of sex  4  Phylogenetics  8  Eukaryote phylogeny....  13  Order of branches at the base of the eukaryotic tree..l4  CHAPTER II  CHAPTER III  CHAPTER IV  Specific questions to be addressed  16  Phylogenetic placement of the oxymonads  22  Introduction  22  Materials and Methods  28  Results  34  Phylogenetic placement of the hypermastigotes  58  Introduction  58  Materials and Methods  64  Results  69  Discussion  92  Overview  92  Oxymonad project  92  Hypermastigote project The ancestral sexual cycle and its infered traits  • •• ill  ....96 101  List of Tables Page#  Table #  Title  Table 1  Guide to taxonomy of some organisms in this thesis  21  Table 2  Predicted fragment size from likely gut symbionts.  56  Table 3  Three classes of clones from PCR replicate 4  57  Table 4  Internal sequencing primers  90  Table 5  Organisms and accession numbers : Chapter Three  91  Table 6  Morphological traits : metamonads and P. lanterna  115  iv  List of Figures Figure #  Title  Figure 1  Effect of sex on variance  Figure 2  "Monophyletic" and "Paraphyletic" diagramatically explained 18  Figure 3  The six to eight kingdom classification of Eukaryotes  19  Figure 4  Basal branches of the Eukaryotic tree  20  Figure 5  Isolation of Pyrsonympha sp. from R. hesperus hindgut  44  Figure 6  PCR products from replicates 1 and 2, Chapter Two  45  Figure 7  Sequence of putative pyrsonymphid  46  Figure 8  Top seven BLAST hits from the putative pyrsonymphid  47  Figure 9  Parsimony tree, Chapter Two  48  Figure 10  Neighbour joining distance tree, Chapter Two  49  Figure 11  Restriction analysis of clones PI 2, P7, P4, P9  50  Figure 12  PCR products from replicates 3 and 4, Chapter Two  51  Figure 13  Restriction digestion of clone from replicate 3, Chapter Two  52  Figure 14  Restriction digestion of clones from replicate 4, Chapter Two  53  Figure 15  BLAST results from sequence of clone L20  54  Figure 16  BLAST results from sequence of clone U39  55  Figure 17  Trichonympha cf. collaris  74  Figure 18  Sexual cycle of Trichonympha from Cryptocercus  75  Figure 19  Evolution of the Parabasalans  76  Figure 20  Isolation of T. cf. collaris from Z. angusticollis  77  Figure 21  :  Page# 17  Trichonympha sp. viewed using phase contrast microscopy  78  Figure 22  PCR products from amplification, Chapter Three  79  Figure 23  Trichonympha cf. collaris ssu rRNA gene sequence  80  Figure 24  BLAST results from Trichonympha cf. collaris sequence  81  Figure 25  T. cf. collaris and R. flavipes gut symbiont 2 ssu rRNA  82  V  Figure 26  Neighbour joining tree, Chapter Three  83  Figure 27  Parsimony tree, Chapter Three  84  Figure 28  Trichonympha cf. collaris probed with universal probe  Figure 29  Trichonympha cf. collaris incubated with the no-probe control 86  Figure 30  T. cf. collaris and S. strix incubated with specific probe  87  Figure 31  Trichomonas and S. strix probed with universal probe  88  Figure 32  S. strix and Trichomonas incubated with the no-probe control 89  Figure 33  Maximum Likelihood tree of E.F.-a sequences  112  Figure 34  Simplified Eukaryotic tree correlated with frequency of sex  113  Figure 35  Hypothetical obligate sexual cycle in a unicellular organism  114  VI  85  Acknowledgements I would like to thank my committee members for their time and good advice. As well, I want to thank the people at Interior Pest Control for providing Reticulitermes specimens.  Thanks also to Ema Chao and Hong  Zhang for providing D N A for my PCR positive controls. There are many additional people around the university that I would like to thank for their help, academic and otherwise. You all know who you are and I am greatful to each of you. I would like to thank Andrew Roger for being the evolutionary postdoc that I never had. I want to thank Tamara Hartson for her drawings of Trichonympha and for making sure that I knew there was always room in front of her fire place and a glass of wine available when I really needed it. I want to thank my friends from the department, especially Patrick Carrier and Andris Maclnns, and from home, especially Ron Odagaki, Chris Eskiw and Ryan McKay. Your support has meant a lot to me. I would like to thank the members of the Redfield lab with whom I worked. Your advice and support has been very important to me over the past three years. Especially I would like to thank Leah Macfadyen for taking the time to proof read my thesis and to Laura Bannister and Shaun Cordes for making sure I reached the end of my thesis with my mental state intact. I want to thank my supervisor, Rosie Redfield, under whose training I have become a much better scientist. And finally I want to thank my family.  vii  Chapter I: Introduction General question : How did sex first evolve? Sexual reproduction is pivotal for so many organisms. Songs are written about it. Salmon swim incredible distances to die for it. Plants develop elaborate coloring and fragrances to encourage it, but this is not so for all of life. Many organisms reproduce effectively and prolifically without sex. Why would this be? What are the advantages of a sexual system? How did sex first evolve and why? Biologists have asked these questions time and time again. I will make my contribution to this debate by addressing the question of when sex originally evolved. Sexual reproduction is costly. In order for it to evolve, an advantage to sex must exist that outweighs its price. This advantage would allow sexual organisms to successfully compete with asexual ones. Some costs (Trivers, 1972), such as males and parental care, are unlikely to have bearing on the origin of sex. However fundamental costs, such as the additional energy and genetic machinery required to participate in sex and the attraction of mates for outcrossing species, are relevant to questions of its origins. A great deal of theoretical work has been done to examine the origin and evolution of sex (Maynard Smith, 1971), (Maynard Smith, 1978), (Bell, 1982), (Hamilton, Axelrod and Tanese, 1990). Models have shown how sex might be advantageous for reproducers with certain traits in given environments. These models are mathematically robust but may not be biologically relevant, as traits and environments are assumed in each model. Although some studies have been done that test specific assumptions, the majority of these assumptions still need to be examined.  1  Rather than testing individual models, I will use phylogenetics to infer traits of the ancestral sexual population. If it is assumed that sex evolved only once, then the divergence that occurred immediately after that can be thought to have produced two lineages. One is the line that gave rise to the majority of the eukaryotes, eventually leading to higher animals and plants. The other lineage diverged away from that line and gave rise to its own modern descendants. This second lineage will be referred to as the deepest diverging sexual lineage and it is this line that I am interested in. I will determine the modern descendants of the deepest diverging sexual lineage. By comparing their sexual cycles with those of more recently diverged organisms, I will infer aspects of the ancestral cycle. Models of the origin of sex will then be re-examined in light of these inferences.  Sex: Definition and Evolution Sex must be defined before its evolution can be discussed. A sexual cycle, in the most elemental sense, can be defined as a cycle of reproduction in which two cells, designated gametes, fuse. A single cell results with double the chromosome content, or ploidy, of each of the individual cells. This process, called syngamy, is followed by the process of meiosis where the chromosome content is reduced back down to its halved state. In most organisms this occurs in two steps. Prior to the first step the D N A content is doubled such that chromosomes have two copies, each referred to as a chromatid. These are connected by a single D N A element called a centromere. The cell now contains two copies of chromosomes, called homologues, each composed of two chromatids. In the first round of meiosis the homologues pair and are segregated to different daughter cells. This division, called the reductional division as the number of chromosomes is halved, is also the division during which molecular recombination, or crossing over, occurs. During the second  2  division the chromatids will separate without D N A replication, producing daughter cells each with a single copy of the chromosomes. This process of two-step meiosis occurs in the vast majority of eukaryotic organisms. However, in some organisms meiosis reportedly occurs in a single step (Cleveland, 1950), (Cleveland, 1950), (Cleveland, 1951). In this process the chromosomes are not duplicated prior to meiosis and segregation of the homologues produces a reduction division yielding daughter cells with half the number of chromosomes as the mother cell, and with no sister-chromatids. There is no known molecular recombination in one-step meiosis. Since these events are seemingly simpler than two-step meiosis and have been reported in some organisms that are presumed to be ancient, it has been speculated that one-step meiosis is the primitive state of the meiotic process (Cleveland, 1950). The details and even existence of one-step meiosis have been challenged however, based on re-interpretation of the microscopic evidence presented by Cleveland (Haig, 1993). Many workers have assumed that the sexual process is monophyletic. This is based on the fact that sex is such a complex process that it is unlikely to have evolved more than once (Cavalier-Smith, 1995), and with the same basic components in all sexual taxa (Maguire, 1992). It has also been noted that the phylogenetic pattern of known sexual versus asexual organisms is most parsimoniously explained by a single evolutionary invention and multiple subsequent losses (Cavalier-Smith, 1995), (Cavalier-Smith, 1981). On the basis of all of these arguments I will also assume that the evolution of sex occurred only once.  3  Theories on the origin of sex Some theories of the evolution of sex have dealt with its maintenance and the protection of a sexual population from invasion by an asexual line. However, four sets of theories have been proposed that are particularly relevant to the origin of sex and therefore to my study. The theory that genetic parasites caused cells to become sexual and those theories dealing with the advantage to a sexual system with respects to beneficial mutations will be summarized first. The last two sets of theories both have a restriction upon them, that they only yield an advantage for sex in a particular population structure. This will be explained and then the theories will be summarized. All four sets of theories will be examined in my Discussion Chapter Four with respect to the inferences I make about the ancestral sexual cycle.  Selfish D N A In 1982, Donal Hickey proposed that a gene "whose sole function is rapid selfreplication", i.e. a selfish gene, could have caused the origin of sex (Hickey, 1982). In his model he shows how an outcrossing sexual population could be invaded by a selfish gene, one that undergoes inter-strand transposition, even if it caused a twofold cost in fitness. Hickey then proposed that a selfish gene could cause sex to originate in an asexual population by coding for non-carrier recognition, cell fusion and transfer of the gene. Bacterial F plasmids, for example, are transmitted in exactly this way. Furthermore it has been shown that the genes responsible for some fungal mating types act by the same transposition mechanism as selfish D N A and have been shown to invade an asexual line (Collins and Gong, 1985). Graham Bell reviewed the evidence supporting this theory in 1993 (Bell, 1993).  4  Freedom from genetic background In asexual organisms a beneficial mutation is tethered to the genome in which it arises. This means that an allele goes to fixation only if its genetic background allows. Sex and recombination provide for the decoupling of alleles thus allowing beneficial ones to spread. The interaction of beneficial mutations with polymorphisms (Strobeck, Maynard Smith and Charlesworth, 1976), or other beneficial mutations (Crow and Kimura, 1965), and even their escape from deleterious mutations (Peck, 1994) have all been modeled to show how successful combinations of alleles might be created by sex. From the point of view of a gene coding for sex, this interaction with beneficial mutations or "genetic hitch-hiking" towards fixation lasts for as long as the linkage does (Barton, 1995). Theories invoking interactions between beneficial mutations as a possible reason for why sex originated have shown it to be effective under both directional (Green and Noakes, 1995) and fluctuating selection (Sasaki and Iwasa, 1987).  Advantage to sex : Generation of rare genotypes? The final two theories share the fact that for them to yield an advantage to sex, a certain distribution of alleles must exist. In these theories, a relative lack of variance in genotypes must exist in a population, such that those near the average are more abundant than would be expected in a random distribution. This could be caused either by negative epistatic interaction between alleles, or else by random drift. In either case because sex causes a randomization of alleles, rare genotypes are reformed (Fig 1). If the rare genotypes are favoured then there is a greater likelihood of a sexual individual having the optimal genotype in a given situation reviewed in (Otto and Michalakis, 1998).  5  Mutation Load reduction The ability to reduce the mutation load of one's descendants could provide an advantage that might allow for the evolution of sex. The mutation load of a genome refers to the reduction in fitness due to the presence of mutations as compared with a reference or mutationless genotype. As an asexual line persists, deleterious mutations accumulate. In 1964, H.J. Muller described the process Muller's Ratchet whereby a random drift event removes the optimal genotype from a population thus making the new optimal genotype more mutationally laden than the last (Muller, 1964). In the case where rare genotypes are under-represented, recombination might allow mutationless alleles to be shuffled into a single individual thus reforming the optimal genotype.  At the same time deleterious  mutations could be shuffled into a different individual which would then be lost from the population. This eliminates several mutations at one time and reduces the overall load of the population. Similarly if, under selection, alleles experience negative epistasis, genotypes with multiple mutations will be less successful than might be expected and therefore occur in lower frequency. The randomization of alleles due to sex will reform both optimal genotypes and highly mutated ones. The former will thrive, the latter will perish. Kondrashov (Kondrashov, 1994) showed that a similar theory would apply to ploidy cycles, which involve meiosis but not true sex. In his model, alleles from a polyploid individual were shuffled down to a haploid state, whereby all genomes but one were randomly destroyed. When the organism returned to its normal polyploid state, there was a probability of the remaining genome being mutationless. In this case the mutational load would be reduced relative to the original polyploid individual. Kondrashov has suggested that ploidy cycles might be a precursor to a true sexual cycle.  6  Defense against parasites Increased defense against parasitic attack is a further advantage of sex (Hamilton, Axelrod and Tanese, 1990), (Otto and Michalakis, 1998). Since parasites act on a population, they would be best adapted to the most common host genotype and therefore rare host genotypes would be less susceptible. Sex would act to generate the rarer genotypes thus producing individuals better able to evade their parasites if these rare genotypes were underrepresented in the population. The rare genotypes might be underrepresented if a recent change in the environment allowed selection for traits previously selected against. In a small population genotypes might be underrepresented due to sampling. Sex could also act to create genotypes not present in the current population. Asexual populations must do this through mutation or migration, both slow processes. However, because sex allows for new combinations, alleles present in other genomes could be recombined to produce genotypes either lost by drift or not yet generated. Should the best adapted genotype for a given situation be absent in a population, it would be more likely to be generated in a sexual population than in an asexual one. The final way in which sex provides for the defense against parasites is by allowing selection to change allele frequencies more rapidly in a sexual population, than in an asexual one. This will only occur in populations that have a lower variance in their genotype frequencies than would be expected by a random distribution. In these cases parasites would have less time to adapt to a given genotype in the sexual population and therefore be less effective. Models that show sex providing a defense against parasites have dealt primarily with the maintenance of sex and the protection from invasion by an  7  asexual mutant. However, the theory may still be applicable to the origin of sex and therefore it has been included here. It should be pointed out that these theories are not mutually exclusive. Sex originating due to a selfish gene does not preclude its being beneficial for any of the other reasons. Similarly decoupling of alleles may be a way of reducing mutational load. There are many other theories for the evolution of sex. However, those covered here are most pertinent to my study as they involve assumptions about the first sexual reproducer or its sexual cycle.  Phylogenetics General introduction Phylogenetics is rife with specialized terminology and methods.  Because  phylogenetic analysis was used extensively in my thesis, I have included an introductory section to the field. Phylogenetics is the study of how organisms are related. Visually, tree diagrams may represent this as the horizontal distance along lines connecting two taxa. In this case the vertical spacing of the phylogenetic tree is irrelevant. Similarly the vertical order of organisms on a tree can be changed, by rotating the branch point connecting two organisms, without changing their relationship. Trees may also be drawn with vertical distance representing relatedness making horizontal distance meaningless. Any group of organisms that encompasses all taxa descended from a common ancestor is called monophyletic (Fig 2A). However, if a group of organisms includes some, but not all of the descendants of an ancestor which is itself part of the taxon, that group is designated paraphyletic (Fig 2B). The term polyphyletic describes a group of organisms deemed related based on characters later determined to be shared due to convergent evolution.  8  By knowing the relationship among organisms, similarities between descendants of a common ancestor can be used to infer the ancestral state of a given characteristic. It is important, when inferring ancestral states, to keep in mind the possibility that similarities might not be ancestral, but due to parallel evolution. This is especially important in cases of traits that are strongly selected for or when the shared characteristic can might be due to loss of the character. Differences in a particular line can also be correlated with proposed selective forces or environmental changes once the ancestral state of a characteristic is tentatively established. The nature of phylogenetics is that a relationship is established using whatever information is available. Provided that the analysis is scientifically robust, that relationship is accepted. However, it is always understood that the relationship can be strengthened or weakened by phylogenetic analysis of other evidence. This is an on-going process and so phylogenetic trees are provisionally accepted and have various degrees of certainty associated with them.  Morphological Phylogenetics Various methods can be used to infer relatedness among organisms. Traditionally, morphological characters were tallied and the number of shared traits used as a measure of relatedness. In 1866 Ernst Haekel produced the first modern phylogenetic tree in this fashion (Haeckel, 1866) and the method has been used ever since. Morphological data can be collected by visual identification of characters and applied to fossils. This allows the state of characters, presumed to be ancestral, to be directly observed. This method is subjective both in the description of characteristics and in the way that characters are weighted when determining relatedness. Therefore it may lead to a more accurate picture of evolution but could also lead to false conclusions  9  should biological importance be mistakenly assigned. Perhaps the greatest danger of using morphological data is that common characteristics may not be shared due to descent but to convergent evolution. Common traits could evolve due to similar environmental factors. Likewise the absence of a trait could be ancestral or due to gain and subsequent loss. Because of the small number of traits used to assign relatedness in morphological studies each trait increases in potential relative importance and each convergent trait has a serious effect.  Molecular Phylogenetics The technology to sequence proteins and nucleic acids has opened up a new field of phylogenetics. Sequences of the same gene can be compared and relatedness mathematically assessed. The premise, first suggested by Pauling and Zuckerkandl (Pauling and Zuckerkandl, 1963), is that the more time elapsed since two lines diverged, the more independent mutations will be accumulated in homologous molecular sequences. By comparing the number of differences in these sequences, relatedness can be determined. Although character states in molecular studies are unambiguously determined and all traits may be weighed equally, there is still a significant amount of judgment used in the alignment of gene sequences. This may have a major effect on the result and so molecular studies are not greatly more objective than morphological ones. Molecular studies tend, however, to use more traits than morphological ones. For a morphological assessment 20 to 30 traits would be considered a large data set. When using molecular data each sequence position is considered to be a trait, and therefore thousands of traits can be compared. This makes the effect of each convergent character in the study less severe. Although the chance of a given homologous position being identical due to convergence is high, when taken  10  together with the many other molecular traits, the effect of each convergent position is low. Convergent structures, that might be morphologically similar, can be coded for by significantly different genetic sequences. There may be variability at the level of genotype that is undetectable at the phenotypic level. Just as morphological data can be misleading, so can molecular data. The position of an organism on a molecular phylogenetic tree is sensitive to the sequence alignment, as well as to the number and choice of taxa included in the study. This can be corrected by careful rechecking of alignments and including a wide representation of taxa in the study. A danger perhaps not as detectable is divergence due to unequal mutation rate, both over time and between different lines during the same time period. Sequence divergence is seen as a measure of relatedness. In the past it has been generally assumed that the rates of change in a given gene are relatively constant. In recent years this assumption is emerging as being less and less true. Some phylogenetic methods take this into account, but it is best countered by acknowledging that phylogenetic conclusions from a single gene can not be taken as definitive. It is important to use both morphological and molecular data when assessing relatedness. Results derived from the two types of data should agree. If they do not, then a reasonable explanation for the difference must be offered. Similarly studies from multiple genes should be consistent in their results.  Molecular Phylogenetic Methods Molecular sequence data can be analyzed by two general methods. Distance methods (Pauling and Zuckerkandl, 1963), such as neighbour joining, compare percent identities between organisms or groups of organisms and calculate this as a distance value. The sequences are then arranged such that the difference between the calculated distance and the distance summed along the branches connecting taxa  11  on the tree are minimized. These techniques are fast and therefore able to deal with many taxa. Discrete character methods (Felsenstein, 1995) use character states at each sequence position to determine the number of evolutionary steps required to produce a given tree. These methods estimate possible arrangements of taxa and determine the tree with the minimum number of changes along each branch. Simple parsimony methods assume that all changes are of equal probability. Maximum likelihood methods assign different probabilities for each type of character change and therefore give a more reliable tree. However, this technique is time consuming, sometimes prohibitively so.  Molecule choice A critical variable of a molecular study is the gene being used for comparison. Different genes change at different rates. For a reliable comparison there must be enough change to distinguish between closely and distantly related taxa, but not so much that the sequence has become randomized. The choice of gene must be tailored to the degree of relatedness in the group of taxa being studied. A great deal of time has elapsed since the divergence of the branches at the base of the eukaryotic tree, and so a conserved gene must be used. There are several candidates e.g. ribosomal RNA, elongation factor a, and R N A polymerase. Small subunit ribosomal RNA ssu rRNA has both advantages and disadvantages. It can be potentially misleading because of its variation in substitution rate within the molecule and between taxa (Neefs et al, 1993). However, there is a large database for small subunit ribosomal R N A sequences, which is important as the number and relatedness of taxa included in a molecular study can have significant effects on its outcome. This, along with its overall  12  conservation and the fact that it exists in multiple copies in most genomes, thus making for easier PCR amplification, caused me to choose this gene for my study.  Eukaryote Phylogeny Phylogenetics is the study of relatedness between taxa and groups of taxa. Classification is the organization of taxa into a hierarchical system based on inferred evolutionary descent. The classification of kingdoms in the eukaryotic domain is the subject of significant debate, with arguments of which groups deserve what rank, and the relative importance of characteristics used for classification, being some of the relevant factors. Because the debate involves a great deal of terminology, a simplified version of the five kingdom eukaryotic tree (Fig 3) will be outlined here. This division of the eukaryotes, based on both morphological and molecular studies is in line with most classification systems, but for a more in-depth discussion see Corliss (Corliss, 1994), Cavalier-Smith (Cavalier-Smith, 1993) and Margulis (Margulis, 1996). Animals, plants and fungi are each placed in their own kingdoms. In trees based on small subunit rRNA, the fungi and the animals are more closely related to each other than either are to the plant kingdom (Cavalier-Smith, 1993). The unicellular eukaryotes can be classified into two kingdoms. Organisms in the kingdom Chromista (Cavalier-Smith, 1981) have compound flagellar hairs, plastids within their endoplasmic reticulum, or are thought to have lost these characters, based on molecular phylogenetic evidence. The majority of chromists are unicellular but some species are colonial or multicellular. The term protozoa can refer to any flagellated phagotrophic unicellular organism. However, classification in the kingdom Protozoa (Cavalier-Smith, 1981) is slightly more strict. Any eukaryote that does not have the requirements to fit in  13  the other four categories is placed in this kingdom. Consequently the kingdom Protozoa is paraphyletic based on both molecular and morphological evidence. Although not currently a full kingdom, the sub-kingdom Archezoa (CavalierSmith, 1983) is of interest as it contains some of the organisms at the base of the eukaryotic tree. This sub-kingdom of the Protozoa houses organisms whose lack of mitochondria, hydrogenosomes, Golgi bodies, and peroxisomes was once thought to be primitive. At its inception, this group included metamonads, microsporidia and archamoebae, as well as the parabasalia. However, in recent years, taxa have been removed from this sub-kingdom based on molecular phylogenetic studies (Clark and Roger, 1995), (Keeling and Doolittle, 1996), or evidence of secondary loss of mitochondria from the group (Germot, Philippe and Le Guyader, 1997), (Roger et al, 1998). Of the six classes listed as archezoan (Corliss, 1994), only the retortomonads and the oxymonads have not been demonstrated to be secondarily simple.  Order of branches at the base of the eukaryotic tree. When addressing the origin of a trait, it is crucial to examine the part of the phylogenetic tree where that trait first appears. In the case of most general eukaryotic characteristics, this is in the branches near or at the base of the phylogenetic tree. Sex is no exception. The removal of taxa from the sub-kingdom Archezoa does not necessarily mean that these taxa are not basal. The diplomonads and trichomonads have both retained their status as basal and primitive despite recent evidence that they diverged after the original symbiosis of mitochondria (Roger, Clark and Doolittle, 1996), (Roger et al, 1998). The order of branches on ssu rRNA trees varies somewhat but the consensus is that the diplomonads are the earliest diverging eukaryotic  14  branch, followed by the microsporidia, trichomonads, percolozoa, and euglenozoa (Fig 4A). However, recent evidence shows that the amitochondriate Microsporidians are likely to be a fungal group whose apparently primitive features and molecular sequence divergence are secondarily derived. This conclusion is primarily based on evidence from mitochondrial heat shock protein sequence that places them with fungal taxa in phylogenetic trees (Germot, Philippe and Le Guyader, 1997) as well as tubulin gene sequences (Keeling and Doolittle, 1996). With the removal of the microsporidia, the order of the basal eukaryotic taxa is the diplomonads followed by the trichomonads, percolozoa and then the euglenozoa (Fig 4B). Of these taxa only the diplomonads have not been reported to have sex, and have been shown to be asexual based on population genetics studies (Tibayrenc et al, 1991). Classification based on morphological characteristics places the diplomonads within the phylum Metamonada (Cavalier-Smith, 1987), (Corliss, 1994). Along with the diplomonads in this phylum are two molecularly uncharacterized classes, the retortomonads and the oxymonads (Fig 4B). I will use the short hand term oxymonads when referring to the class Oxymonadea. The retortomonads are not known to be sexual but the oxymonads have clearly defined and characterized sexual cycles (Cleveland, 1950), (Cleveland, 1950), (Cleveland, 1950), (Hollande and Caruette-Valentin, 1970). The trichomonads are also a class within a larger morphological phylum. The phylum Parabasalia (Corliss, 1994) contains two classes, the trichomonads and the hypermastigotes (Fig 4B). The ssu rRNA of several trichomonads have been sequenced but to this point no hypermastigote sequences have been identified. With the exception of Mixotricha paradoxa for which there is no sequence data available, all known trichomonads are asexual, whereas the hypermastigotes have well characterized sexual cycles (Cleveland, 1956).  15  Table 1 shows the groups included in the metamonads, parabasalans and in the umbrella group that I have designated "Higher Organisms". This table will be useful for reference during my justification for the specific questions to be addressed in the thesis as well as the discussion of the results from phylogenetic studies.  Specific questions to be addressed In order to establish the first sexual branch to have diverged from the eukaryotic tree, I will address the following questions.  1) What is the phylogenetic placement of the sexual archezoan class, the oxymonads, based on small subunit ribosomal RNA sequence?  2) What is the phylogenetic placement of a sexual protozoan class, the hypermastigotes, based on small subunit ribosomal RNA sequence?  By determining the placement of these two classes based on molecular sequence data, I will corroborate or provide conflicting evidence for their morphological classification. This will help to clarify the organization of the branches at the base of the eukaryotic tree and indicate where on the tree sex first appeared. The ancestral sexual cycle will be inferred from common features of the modern ones. This may give insight into the validity of the current mathematical models describing the origin and evolution of sex.  16  17  Figure 1 : Effect of sex on variance. The X-axis may represent number of mutations or genotype  B 18  Figure 2: "Monophyletic" and "Paraphyletic" diagramatically explained A=Here A, B, C and D are monophyletic as the group includes all descendants of their most recent common ancestor B=Here C and D are paraphyletic as this group excludes some of the descendants of the most recent common ancestor.  A  Figure 3 : Simplified diagram of a five kingdom classification of the Eukaryotes. Two protozoan branches are shown to indicate that the kingdom Protozoa is not monophyletic 19  A) Tree based on ssu rRNA evidence B) Tree based on ssu rRNA as well as HSP 60 and morphological evidence. Bacteria are included as an outgroup. "Higher" organisms are defined in Table! 20  Figure 4: Basal branches of the Eukaryotic tree.  Bacteria  Me  Parabasalia  Bacteria  Trichomonads  ypermastigotes?^  Percolozoa  Diplomonads 1 ' , Oxymonads?/ tamonads  B Higher  Diplomonads (Metamonads)  Microsporidia  Trichomonads (Parabasalia)  Percolozoa  Euglenozoa  Higher  euglenozoans Animals (kingdom) Plants (kingdom) Fungi (kingdom) Chromists (kingdom)  percolozoans  trichomonads (class)  hypermastigotes (class)  Trichomonas Trichomitus Psalteriomonas Naegleria Valkampfia Euglena Mnemiopsis Cyanophora Schizosacaromyces Ochromonas  H ex ami t a Retortomonas Oxymonas Pyrsonympha Trichonympha collaris campanula sphaerica agilis vaginalis trypanoides lanterna  21  Upper level taxons are umbrella groups used when describing trees. Within each group are several subgroupings. A n example genus is given for each subgroup, and species names are given if the species is specifically refered to in the thesis. This table is meant for readers unfamiliar with the taxomonic classification of lower eukaryotes and as a clarification of the term "Higher" organisms.  Table 1: Guide to taxonomy of some organisms in this thesis  "Higher organisms"  Parabasalia  retortomonads (class) oxymonads (class)  lamblia ardeae inflata  diplomonads (class)  Metamonads  Giardia  species  Common name of groups Example genus in the Taxon  Upper level Taxon  Chapter II: Phylogenetic placement of oxymonads Introduction In my search for candidate descendants of the deepest diverging sexual lineage, I examined the eukaryotic tree for the branch where sex first appears. The diplomonads appear to be the deepest diverging eukaryotic lineage and are asexual (Tibayrenc et al, 1991). All branches above them are reported to have sexual representatives (Margulis et al, 1993). The origin of sex, therefore, is likely to have occurred near the time of this divergence. As listed in Table 1, the diplomonads are classified in the phylum Metamonada metamonads on the basis of morphology, of which the retortomonads and the oxymonads are also part (Cavalier-Smith, 1987), (Corliss, 1994). Although there is no known sexual cycle in retortomonads, the oxymonads have sexual cycles well characterized by light (Cleveland, 1950), (Cleveland, 1950), (Cleveland, 1950) and, in some cases, electron microscopy (Hollande and Caruette-Valentin, 1970). For this reason the oxymonads are strong candidates as descendants of the deepest sexual lineage.  Classification and Morphology of the oxymonads The descriptions of general oxymonads, and of Pyrsonympha in particular, have interspersed in them explanations of how the morphology has been used to make phylogenetic inferences. This will serve both to clarify the morphological phylogenetic placement of the oxymonads and also to reinforce the fact that classifications are not only of sequences but of the organisms to which they belong. The classification of the oxymonads is controversial. Dyer emphasized the pyrsonymphids with the classification of the oxymonads in the class Pyrsonymphida, containing a single order (Dyer, 1993). The Illustrated Guide to the Protozoa (Lee, Hutner and Bovee, 1985) reduces the oxymonads to order status but 22  designates several families, including the Oxymonadidae and Pyrsonymphidae. Cavalier-Smith (Cavalier-Smith and Chao, 1996), considered the oxymonads to be a class (Oxymonadea) with two orders, the Oxymonadida and the Pyrsonymphida. I agree that the oxymonads are morphologically different enough from the other lower protists to deserve class status and that the distinction between the Pyrsonymphida and the Oxymonadida as orders based on their morphological distinctiveness is also justified. I will therefore adopt the classification of CavalierSmith. The oxymonads are amitochondriate symbionts of termites and wood-eating cockroaches. On average they are 100 |im long but range in size from 5 u.m for the smallest Oxymonas species (Cross, 1941) to 320 (im for the "giant" forms of Streblomastix. (Kidder, 1929). Flagellar numbers of 4, 8 or 12 are characteristic of this group. Some have symbiotic bacteria and many have spirochetes attached to their exterior. The occurrence of a motile (usually) contracting axostyle is the chief classifying characteristic of this class, with anterior nuclei, a cytoskeletal holdfast or rostrum organelle and a pre-axostylar apparatus also being characteristic.  Order Oxymonadida The Oxymonadida include many genera (Lee, Hutner and Bovee, 1985). The ones of special importance to the evolution of sex are : Oxymonas, Notila, and Saccinobaculus. Shared characteristics of these organisms are four anterior flagella, and a holdfast or rostellum apparatus used for attachment to the host gut lining.  Order Pyrsonymphida The Pyrsonymphida include Pyrsonympha/Dinenympha and Streblomastix. There has been debate as to whether Pyrsonympha and Dinenympha are morphs of the same organism or two separate genera. Because the morphological descriptions  23  differentiate Pyrsonympha and Dinenympha, I will keep the designation used in the original reference. However, based on work by Hollande (Hollande and CaruetteValentin, 1970), I will assume that the organisms are the same. I caution that this assumption warrants further examination. Dinenympha individuals may be as small as 40 |im, and Pyrsonympha as large as 170 um have been reported (Grasse, 1952). A holdfast organelle extends slightly off center and anteriorly from Pyrsonympha attached to the gut lining of their hosts (Cochrane et al, 1979). This holdfast is an extension of the mastigont system, an organizing center for flagella and other cytoskeletal elements in the cell. The flagellar basal bodies are organized with all four clustered at the anterior end of the cell. This basal body arrangement is one of the primary features upon which the metamonads are unified. Between the basal bodies lies the pre-axostylar ribbon, a structure of intercrossed fibers. From this ribbon is derived a paracrystalline cytoskeletal array (thousands of microtubules) organized into rows called the axostyle (Brugerolle, 1991). This flexible structure runs the length of the cell, with the posterior end of the axostyle sitting free in the cytoplasm. The pre-axostylar ribbon and axostyle are the strongest morphological features binding the oxymonads together. Both four and eight flagella have been described running recurrently in a left handed spiral along the surface of Pyrsonympha. Smith (Smith, Stamler and Buhse, 1975) reports that the flagella are attached to an undulating membrane system beneath which vesicles, membrane invaginations, and cytoplasmic granules can be seen. The nucleus, containing one to three nucleoli, is normally in the anterior part of the cell. The cell surface is covered with spirochetes, that may attach via tubular scales (Smith, Stamler and Buhse, 1975). Additional symbiotic bacteria are known both in the cytoplasm of Pyrsonympha and possibly within the 20-30 (im nucleus itself  24  (Bloodgood et al, 1974). The granular cytoplasm of Pyrsonympha contains large food vacuoles with wood, bacteria and "unidentified debris". Rough endoplasmic reticulum and "membrane bound, densely staining bodies" were also reported by Bloodgood (Bloodgood et al, 1974). Bloodgood speculated that these membrane bound bodies may be evidence of a Golgi body (Bloodgood et al, 1974) and CavalierSmith proposed that they might be hydrogenosomes, or peroxisomes (CavalierSmith, 1987).  Sexual and ploidy cycles of oxymonads A description of the oxymonad sexual and ploidy cycles will be given here and used in Chapter Four when discussing the evolution of the oxymonads and their relevance to the origin of sex.  Order Oxymonadida The sexual cycle of Oxymonas (Cleveland, 1950) is representative of organisms in the Order Oxymonadida. Oxymonas starts its sexual cycle as a haploid cell which undergoes chromosome replication to become a gametocyte. Morphologically indistinguishable from a mitotic cell, the gametocyte divides to produce two gametes. These fuse to form a zygote which then undergoes a one-step meiosis. The axostyle and flagella are discarded and regenerated several times during this process.  Order Pyrsonymphida Pyrsonympha does not have a true sexual cycle with meiosis and gamete fusion. Rather, it undergoes a ploidy cycle involving repeated rounds of meiosis to a reduced ploidy state followed by rounds of mitosis without nuclear division to regain polyploidy. This cycle is triggered by starvation and subsequent molting of  25  the termite host. Hollande (Hollande and Caruette-Valentin, 1970) induced the ploidy cycle of Pyrsonympha by starving the termite host. After 10 days, the attached Pyrsonympha dropped off the gut lining and into the hindgut cavity. Cells appeared ruffled and free of food vacuoles. Five days later, cells became spherical and opalescent, having lost their rostrum. The axostyle was severed from the mastigont system with only the most anterior portions of the axostyle remaining connected, similar to the discarding of the axostyle in other oxymonad sexual cycles. The cells divided twice over the subsequent six hours without doubling of the flagella. Using electron microscopy, Hollande found synaptonemal complexes in the nuclei of dividing cells. He also made D N A measurements that demonstrated a halving of the D N A content of the cells in each division. Together these observations suggested that the divisions were meiotic in nature.  Keticuliterm.es hesperus/flavipes gut fauna As mentioned earlier, oxymonads are gut symbionts of termites and woodeating cockroaches. One of the best characterized oxymonad containing systems is the gut fauna of the subterranean termites, Reticulitermes flavipes (eastern species) and R. hesperus (western species). In a 1995 study, D N A from 25 R. flavipes hindguts and their contents was extracted and the ssu rRNA gene was amplified using universal eukaryotic primers (Gunderson et al, 1995). Two unidentified parabasalan sequences were obtained and placed in GenBank. These sequences were never matched to organisms. In a census of the R.flavipes.hindgut community, oxymonads composed 7075% of the identified protists counted (Grosovsky and Margulis, 1982). R. hesperus, endemic to the interior of British Columbia, is known to contain Pyrsonympha major and two species of Dinenympha (Kirby, 1934).  26  Due to the availability of the termites and the well characterized nature of the system, I decided to determine the phylogenetic placement of Pyrsonympha from R. hesperus using the ssu rRNA gene.  Why oxymonads are candidate descendants of the first sexual lineage Although some details of the processes may be questionable see discussion in Chapter Four, Cleveland's work (Cleveland, 1950), (Cleveland, 1950), (Cleveland, 1950) clearly shows that the oxymonads are sexual. Currently the oxymonads are placed, based on morphological similarities, in the phylum Metamonada with the diplomonads (Fig 4B). This classification is based on a characteristic arrangement of basal bodies (Brugerolle, 1991) and a lack of plastids, mitochondria, peroxisomes, hydrogenosomes and Golgi bodies (Cavalier-Smith, 1987). Due to their amitochondriate status, their lack of Golgi bodies, and the prokaryotic 16s-like nature of their ribosomes (Leipe et al, 1993), (Van Keulen et al, 1993), the metamonads are assumed to be ancient. This has been confirmed for the diplomonads by ssu rRNA sequence data (Van Keulen et al, 1993), (Cavalier-Smith and Chao, 1996). If placed as a sister group to the diplomonads, the oxymonads would be descendants of the deepest diverging sexual line. However, their relation to the metamonads seems primarily based on the lack of characteristic organelles.  Because  these losses may have been secondary, it is important to test their phylogenetic placement with molecular data.  27  Materials and Methods Finding, identifying and keeping termites Western Subterranean termites R. hesperus were identified in a woodpile in Kelowna, B.C. by employees of Interior Pest Control. A sample of wood containing the termites, along with a quantity of soil, was collected and kept in a glass aquarium. Aluminum foil was attached over the edges of the aquarium to prevent the termites from eating the sealant and escaping. A Plexiglas lid with wire mesh covering a large air hole was constructed and weighed down to further discourage escape. Water was sprayed into the tank every 2-3 days to maintain moisture.  Dissection and identification of Pyrsonympha The gut contents of a single R. hesperus individual were emptied into a solution of modified Trager's medium (Buhse, Stamler and Smith, 1975). Approximately 20 ul of this solution was diluted in 500 ui of fresh medium and observed under magnification not recorded using an inverted microscope (PhotoZoom™ Inverted Microscope, Bausch + Lomb). Pyrsonympha individuals were identified based on their shape. The lower half of their bodies were twisted in a left handed helix, while the upper half was not twisted but ended in a knob-like structure. They were also identifiable by their size. By far the largest protist in the gut fauna, and therefore easily identifiable, Pyrsonympha of various sizes (not recorded) were observed and collected. Individuals were drawn into a blown glass micro-pipette, rediluted into fresh medium (500 |il), and reselected (Fig 5), with 50-75 individuals being collected in a 1.5 ml microfuge tube.  D N A extraction 28  Cells were pelleted by centrifugation at 2000-3000 rpm for one minute. A l l other centrifugations in this thesis were done at 13000 rpm. After aspiration of the supernatant, the cells were macerated with 50 ul of cell lysis buffer (0.1 M Tris p H 8, 1.4 M NaCl, 0.04 M EDTA, 2% w / v CetylTrimethylAmmonium Bromide), heated to 65 °C, and incubated with an additional 150 ul of cell lysis buffer for 15-30 minutes. Phenol/chloroform (200 ul) and distilled water (200 ul) were then added. The mixture was vortexed and centrifuged at 4°C for 10-15 minutes. The aqueous layer was further extracted with 200 ul of chloroform, and centrifuged at 4°C for 10 minutes. The aqueous layer from this step was incubated overnight at -20°C with 2.5 volumes of 95% ethanol. This was then centrifuged for 20 minutes at 4°C, washed (200 ul 70% ethanol, centrifuged 4°C, 10 minutes) and dried.  PCR With the help of Dr. Cavalier-Smith, a 5' primer ( T G A A A C T T A A A G G A A T T G A C G G A ) was designed by aligning various small subunit rRNA sequences and choosing a region for amplification conserved in both diplomonad sequences, thought to be the closest sequenced relatives to oxymonads, and higher organisms. This primer, designated 5' NEWR, corresponds to bases 921943 of the Hexamita inflata ssu rRNA sequence and was used in coordination with the 3' primer of Gunderson et al (Gunderson et ah, 1995). A l l primers were synthesized by the Nucleic Acid and Protein Synthesis Unit UBC using ABI 380B, 394 and 391 Oligonucleotide synthesizers. Primers were used at ImM in a total volume of 50 ul per reaction with lx Boeringher Mannheim PCR buffer + Mg2+. dNTPs were used at a concentration of 0.2mM.  29  The entire sample of Pyrsonympha D N A was used for a single P C R reaction. Cercomonas sp. D N A (50 ng), donated by Ema Chao, was used in a positive control reaction. Water from the Milli-RO 60 Reverse Osmosis system / M i l l i - Q Plus Water Systems was used as a n o - D N A control, and A R T 200 aerosol tips (Continental Laboratory Products) were used for all micropipetting involved in the P C R reactions. Each reaction was set up on ice in a 500 |i.l PCR tube and cycled in a Perkin Elmer Cetus D N A Thermal Cycler 480. Reactions began with an initial cycle of 95 °C for one minute, followed by 1 minute of 45 °C and 3 minutes of 72 °C. This cycle was repeated an additional 28 times with the initial heating step running at 94 °C for 10 seconds.  A final cycle of 94 °C for 10 seconds, 1 minute at 45 °C and 4 minutes at  72 °C was done to promote complete synthesis of all copies.  PCR product extraction and purification PCR products were purified by gel electrophoresis on T A E gels (0.04 M Trisacetate, 0.001 M E D T A ) with 0.8% agarose. Gels were run submerged in T A E with ethidium bromide in Model BI gel boxes (Owl Scientific Plastics INC) and using a FB 105 (Fisher Biotech electrophoresis systems) power supply. P C R products were observed under ultraviolet light and cut out of the gel. These were extracted and purified using the protocol and materials from a Gene Clean II kit (BiolOl). The concentration of the final purified product was estimated against Lambda Hindlll ladder (Gibco BRL) and l k b ladder (Gibco BRL) size standards. Cloning P C R products were ligated into p G E M - T vectors using all materials and protocols from the p G E M - T cloning kit (Promega) at an insert to vector ratio of 6:1. Plasmids were then transformed into E. coli DH5oc cells using a standard calcium chloride protocol. In addition to the control insert provided by the p G E M - T kit, and  30  a transformation control with 10 ng of pGEM 7+ D N A (Promega), water was substituted for D N A as both ligation and transformation no-DNA controls.  Clone selection Transformants were blue/white screened on 100 ml LB plates with 100 ug/ml ampicillin (Sigma) and 100 ul of 20 mg/ml X-gal (US Biological). White clones were restreaked and grown overnight at 37°C. Clones that were white after restreaking were grown in 5 ml of LB Amp (100 ug/ml) broth with rolling at 37°C overnight. An aliquot of these cells was frozen at -70°C with 10% glycerol. Plasmid D N A was prepared from a second aliquot of this culture. At room temperature, 3 ml of cells were centrifuged for 2 minutes. Pelleted cells were then resuspended in 100 ul of LB and incubated on ice for 5 minutes with 200 ml of Solution 2 (0.2N NaOH, 1% SDS). This was further incubated for 5 minutes at room temperature with 150 ul of Solution III (3M KOAc, 11.5% acetic acid). After extraction with 450 ul of phenol/chloroform, the solution was mixed by vortex for 10 seconds and centrifuged at room temperature for 3 minutes. The aqueous layer was then incubated for 5-15 minutes at room temperature or -20°C with 1.0 ml of 95% ethanol. The precipitated plasmid D N A was collected by centrifugation for 10 minutes, and washed with 500 ul of 70% ethanol. After 2 minutes of centrifugation the pellets were dried at 65 °C and resuspended in distilled water. Plasmids were linearized at a site 14 bp from the origin, using Apal (Pharmacia Biotech). Inserts were removed from the vector using the Apal site and a Pstl (Gibco BRL) site 22 bases after the insertion site. To test for a characteristic putative pyrsonymphid restriction pattern, two digests were designed such that the enzymes cut at sites in the putative pyrsonymphid ssu rRNA sequence but not in that of Neobulgaria, the sequence that my initial BLAST search identified as matching with the smallest p value.  31  Plasmids tested for the characteristic putative pyrsonymphid restriction pattern were cut with BanI (New England Biolabs) and Rsal (Pharmacia Biotech) in separate digests at 37°C for 90 minutes. Enzymes were used at 0.5 ul in 20 ul reactions with OPA buffer (Pharmacia Biotech) used as directed (Pharmacia Biotech Catalogue, 1995). Plasmid D N A was further purified to remove residual RNA before being submitted for sequencing. After plasmid DNA was prepared as stated in the previous section, it was dissolved in 100 ul of TE (lOmM Tris.HCl(pH8.0), I m M EDTA(pH8.0)). 300 ul of 4 M lithium chloride was then added and the mixture was incubated for 30 minutes on ice. The precipitated RNA was removed by centrifugation for 5 minutes and the supernatant was incubated with 600 ul of isopropanol for 10 minutes at room temperature. The precipitated D N A was collected by centrifugation for 10 minutes, resuspended in 100 ul of TE and incubated with RNAse at 37 °C for 1 to 2 hours. To remove the RNAse, the solution was mixed with 100 ul of phenol/chloroform, and centrifuged for 10 minutes. The aqueous phase was further extracted with 100 ul of chloroform and spun for 15 minutes. The plasmid D N A was precipitated by incubation with 1/10 volume of 3 M sodium acetate and 2.5 volumes of 95% ethanol. This aqueous solution was incubated at -20 °C overnight. The DNA was collected by centrifugation for 20 minutes and washed with 200 ul of 70% ethanol. The washed pellet of D N A was resuspended in 20 ul of distilled water for every 1.5 ml of initial cell culture used.  Sequencing Each sequencing run required 500 ng of plasmid DNA. Plasmids were sequenced from the M13 forward and reverse sequencing sites of pGEM-T. Primers were provided and sequencing was performed by the NAPS unit (Biotechnology Laboratory, U.B.C.) using an ABI 373 or 377 DNA sequencer.  32  Since the insert sequence began with the primer, the Amplify 1.2 PCR simulator (Engels, 1992) was used to find the primer sequence within the sequence text file. Once insert sequence was found, it was checked by eye against the sequencing read-out for mis-called bases. This established the length of reliable nucleotide data. To assemble contiguous regions, sequences were aligned using the Clustal W program (Thompson, Higgins and Gibson, 1994). Discrepancies between bases were checked by eye against the two sequence read-outs.  Phylogenetics Related sequences to the ones obtained in my study were initially found by searching GenBank using the default settings at the Basic BLAST database search page (http://www.ncbi.nlm.nih.gov/BLAST/). The putative pyrsonymphid sequence was aligned by eye to an alignment provided by Tom Cavalier-Smith, and visualized using the Genetic Data Environment (Smith et al, 1994). Sequences from this alignment, except that of the putative pyrsonymphid, are available from either the GenBank or EMBL databases. PHYLIP (Felsenstein, 1995) was used to do all phylogenetic analyses. The alignment was bootstrapped 100 times and used with the default parameters in Dnapars and Dnadist with neighbour joining. Any parameter settings not set at default values will be stated on the figures. Phylogenies were derived from complete gene sequences except for the putative pyrsonymphid sequence.  33  Results Initial Pyrsonympha sequence Identification of Pyrsonympha and DNA extraction In order to obtain the ssu rRNA gene sequence from Pyrsonympha a pure sample of these organisms was needed. Since Pyrsonympha is difficult to culture (Dyer, 1993), (Buhse, Stamler and Smith, 1975) it was necessary to manually isolate and collect Pyrsonympha samples. Occasionally during the isolation, other large protists (Trichonympha agilis) would enter the micropipette and these were rejected by discarding that sample of dissection fluid. However, the field of view under which I performed the identification was small and would blur when large amounts of liquid were being taken. It is therefore possible that other gut symbionts may have been collected in addition to the 50-75 Pyrsonympha individuals.  PCR replicates 1 and 2 In order to amplify the ssu rRNA gene from Pyrsonympha, the extracted D N A was subjected to PCR. Based on predicted sizes from the simulated amplification of ssu rRNA sequences using the 5' NEWR and 3' B primers, a product of between 600 and 700 base pairs was expected. Due to the presence of a single band in both the positive control and the Pyrsonympha D N A as well as the lack of bands in the no-DNA control lane, PCR replicate 1 was deemed successful (Fig 6A). This PCR was repeated (PCR replicate 2) using Pyrsonympha D N A collected at the same time as that used in PCR replicate 1. PCR replicate 2 ensured reproducibility (Fig 6B).  34  D N A purification and cloning from PCR replicate 1 To allow for sequencing of the PCR product it was purified, cloned into a vector and transformed into E. coli strain DH5a. Six clones containing the plasmid were recovered. Of these, four clones (designated P4, P7, P9, P12) had inserts of the appropriate size.  Sequencing In order to obtain the identity of the cloned insert, clone PI 2 was sequenced from the M13 forward and reverse primers. Sequence obtained was assembled into a contiguous strand 640 bp long, excluding the primer (Fig 7). This sequence will hereafter be referred to as the putative pyrsonymphid sequence.  BLAST Results In order to determine whether identified sequences present in GenBank were closely related to the putative pyrsonymphid, a series of BLAST queries were submitted. The reliability of a match in a BLAST search is measured by its p value, the probability that a given match would occur in two random sequences. The smaller the p value, the more reliable is the match. Matches to ribosomal R N A sequences from plants, animals, and protozoa were found, all with similar p values (Fig 8). It was necessary to compare the p values obtained from my matches to those calculated for a match of a sequence query to itself and closely related taxa. To do this a 444 bp region of the ssu rRNA gene from Neurospora was sent for BLAST analysis. This region was homologous to a query submitted for the putative pyrsonymphid sequence. Whereas the putative pyrsonymphid query returned matches to animals, plants and fungi with p values of 10~77 to 10"75, the homologous region from Neurospora returned a hit to itself and closely related fungi with p values of IQ'178,  35  Since queries from the putative pyrsonymphid sequence returned matches from all branches of the eukaryotic tree, with equivalent p values, I conclude that the putative pyrsonymphid sequence is derived from a eukaryotic ssu rRNA gene but is not closely related to any sequences present in GenBank at the time of the analysis.  Phylogeny Since no sequences were found to be closely related to the putative pyrsonymphid sequence, phylogenetic trees were constructed using both parsimony (Fig 9) and distance methods (Fig 10) in an attempt to determine the placement of the putative pyrsonymphid sequence relative to a variety of eukaryotic ssu rRNA gene sequences. Bootstrap values for the placement of the putative pyrsonymphid sequence were not high enough to make a judgment about its membership in a particular clade. However, in trees constructed using both phylogenetic methods, the parabasalans and diplomonads formed monophyletic clades with 100 percent bootstrap values and were excluded from the other sequences also with 100 percent bootstrap values (Fig 9, 10). Although no solid placement was established for the putative pyrsonymphid sequence, in no case did it diverge below the parabasalans or as a sister to the diplomonads.  Identity of the putative pyrsonymphid sequence The washing step used to isolate Pyrsonympha was not sufficient to exclude the possibility of contamination. It was therefore necessary to eliminate alternate sources for the putative pyrsonymphid sequence. If this could be done, then the most parsimonious explanation would be that the sequence is derived from the Pyrsonympha targeted in the isolation.  36  Characterization of additional clones from PCR replicate 1 Although the four clones taken from the cloning reaction have the same sized insert, that does not mean that they contain the same sequence, as most ssu rRNA sequences should yield PCR products of between 600-700 bp using these primers. Since only one clone was characterized from PCR replicate 1, it is possible the putative pyrsonymphid sequence is not derived from the targeted  Pyrsonympha  but actually from a rare contaminant in the PCR reaction. This contaminant would represent only a small fraction of the clones in the pool, and could have been chosen by a sampling error. It could have been derived either from a gut protozoan or from a eukaryote not present in the termite gut fauna. To eliminate this possibility, restriction digests were carried out on the three remaining clones obtained from PCR replicate 1. The probability of two unrelated clones having the same restriction pattern is small. However other factors, such as the fact that the sequences being resolved are related (the same gene from different organisms) or the presence of unresolved fragments of different sizes, increase the likelihood of similar RFLP patterns. None the less one predicts that, if the putative pyrsonymphid clone is a rare contaminant, it is probable that other restriction patterns will appear in analyses of the other clones in the pool. Since the restriction digestion patterns of the four plasmids are indistinguishable (Fig 11), I conclude that the clones are derived from the same source. The clone containing the putative pyrsonymphid sequence does not merely represent a minor fraction of the clones from PCR replicate 1, and therefore it is not likely to have been derived from a rare contaminant.  Clones from PCR replicates 3 and 4 Two concerns still remained about the putative pyrsonymphid sequence. It was possible that a contaminant from outside the termite gut fauna and specific to  37  the P C R tube in replicate 1 had swamped the Pyrsonympha D N A . If so, the restriction pattern seen in the putative pyrsonymphid clone should not be present in a pool of clones from an independent P C R reaction. A n additional concern was that, since only four clones were analyzed from P C R replicate 1, a representative picture of the population of P C R products was not seen. This might mean that the probability of the putative pyrsonymphid sequence being derived from the organism targeted appeared greater than it was. If the isolation procedure had been successful then clones with the restriction patterns of the other common gut organisms should not be present in a pool of clones from the P C R reactions. The predicted restriction fragments from p G E M - T with the P C R products from both orientations of the putative pyrsonymphid sequence from P C R replicate 1 and the two unidentified trichomonads sequences derived from symbionts of a Reticulitermes gut, were calculated and are displayed in Table 2. The issues of contaminants specific to PCR replicate 1 and the search for clones derived from other gut organisms can both be addressed by establishing a proper representation of the products amplified in independent P C R reactions.  PCR replicates 3 and 4 In order to get clones from an independent P C R pool, a replicate of the Pyrsonympha P C R experiment was performed (replicate 3). However, because only one clone was obtained from this experiment (see below) a re-amplification of the D N A from P C R replicate 3 was done. The D N A used in these replicates was extracted at the same time and from the same termite gut as that used to obtain the first sequence. Due to the presence of bands of approximately 600-700 bp in both the positive control and the Pyrsonympha D N A as well as the lack of bands in the n o - D N A control lane, the P C R was deemed successful (Fig 12). However, the presence of an  38  additional band of similar size to the previous bands in the Pyrsonympha D N A lanes indicated that additional sequences had been amplified. This ^reproducibility in the PCR prompted me to do separate D N A preparations and clonings to identify the different bands.  Cloning Because two bands of approximately 600-700 bp were seen in the Pyrsonympha D N A lane in both PCR replicates 3 and 4, separate ligations were done to identify the PCR products from each one. Ligations from products of PCR replicate 3 were transformed into freshly competent E. coli DH5cc cells. A single clone from the upper band was retrieved. Ligations from PCR replicate 4 were transformed into E. coli DH5oc cells. These ligations yielded 24 clones (designated U#) from the upper band and 18 clones (designated L#) from the lower band. Clones broke down into two classes based on insert size: those with an insert of 700bp and those with one of 660bp.  Restriction analysis In order to establish the composition of the products from PCR replicates 3 and 4, a characteristic pattern of restriction fragments was established for each clone using BanI and Rsal. The single clone from PCR replicate 3 has a pattern consistent with a putative pyrsonymphid insert, but in the opposite orientation to the first 4 clones (Fig 13). The clones from PCR replicate 4 break down into three classes. The first class contains clones with patterns indistinguishable from that of the putative pyrsonymphid. Clones from the other classes have distinct BanI and Rsal restriction patterns (Fig 14). The breakdown of clones in each class is seen in Table 3.  39  Sequencing of clones from PCR replicate 4 The presence of two classes of clones, in addition to the one previously sequenced, meant that additional sequences had been amplified. In order to identify these, sequence was obtained from a clone from each of classes 2 and 3.  BLAST results, class 2 (clone L20) Two B L A S T queries for this sequence were sent to GenBank. The first query of 121 bp found matches to bacterial 16s ribosomal R N A sequences, with Arthrobacter, an unidentified bacterial sequence and Pseudomonas species yielding p values on the order of 10"39 to 10"37 (Fig 15A). A second query with the entire 390 bp obtained from the insert in clone L20 was also submitted. Again this returned Arthrobacter, the unidentified bacterial species, and Pseudomonas sequences. P values for these matches are on the order of IO"  1 4 2  to I O "  1 2 0  Fig (15B).  Based on the strong matches from the B L A S T searches, I conclude that the insert in clone L20 is derived from a bacterial contaminant. Since the rest of the clones in class 2 had the same restriction pattern as L20, it can be assumed that all of the clones in this class are also derived from bacterial contaminants. These bacteria could have been derived either from the gut fauna or from the soil in which the termite was living.  class 3 (clone U39) The entire sequence obtained from the insert in clone U39, as well as several separate regions of the sequence, were submitted to GenBank for B L A S T analysis.  40  The first 123 bp of the sequence returned algal, plant and fungal matches with p values in the 10"  4u  range. Algal (stiamenopiles) sequences made up 19 of the first  20 matches returned (Fig 16A) with 14 sequences from the algal class Chrysophyceae among them. The middle l l l b p returned matches to four algal sequences, all from the Chrysophyceae with p values of 10"24. Matches to fungal sequences were then returned with p values of lO'^l (Fig 16B). A BLAST query of the entire sequence obtained from clone U39 (462bp) returned 39 stramenopiles sequences with p values ranging from 10~141 to 10~^7. Twenty two of the first 23 sequences were from the class Chrysophyceae, and the first eight were from the Synurales order. The first five were from the genus Mallomonas (Fig 16C). Due to the number of Chrysophyceae or golden algae sequences matched, there is a very high probability that the insert in clone U39 is derived from an organism belonging to this order. However, due to the return of fungal sequences to some BLAST queries without significant drops in the p values, there is still some doubt as to the identity of the sequence.  Percent identity studies for class 2 To address the identity of the U39 sequence in another way, percent identity of the insert sequence from clone U39 was determined with respect to the golden alga Mallomonas, the algal species most strongly returned as a match. The percent identity of the insert sequence from U39 to Neurospora, a representative fungal sequence, was also determined. Since the putative pyrsonymphid sequence would not be in GenBank, the percent identity of that sequence to the insert sequence from U39 was also calculated.  41  The U39 and Mallomonas sequences were 94% identical. The U39 and Neurospora sequences gave only 82.9% identity, similar to the value obtained for Mallomonas when compared with Neurospora (83.3%). The percent identity between U39 and the putative pyrsonymphid was 69.9% and 72.0% between the putative pyrsonymphid sequence and Mallomonas. This indicates that the insert sequence from clone U39 is not closely related to that from the putative pyrsonymphid. From both the B L A S T and the percent identity results, I conclude that the insert sequence in clone U39 is derived from a member of the golden algae, most likely from the Synurales family. The golden algae are small (approximately 10 um) and are present in fresh water. They have very little morphological similarity to the oxymonads. Although it is possible that the insert sequence in clone U39 is an oxymonad and that the oxymonads are actually algae, it is much more likely that a fresh water alga was amplified as a contaminant foreign to the termite gut fauna.  Conclusions from analysis of clones from independent PCR replicates 3 and 4. Determining the representation of various clones from independent P C R amplifications serves to eliminate several possibilities about the identity of the putative pyrsonymphid sequence. Since clones with the identical restriction pattern, or that of the other orientation, to the putative pyrsonymphid sequence were found in the additional independent cloning experiments, the possibility that the putative pyrsonymphid sequence is derived from a contaminant specific to P C R replicate 1 is significantly reduced.  42  Another possibility to be tested was that clones with the insert from the putative pyrsonymphid sequence represented a fraction of the entire population of PCR products. This might mean that the putative pyrsonymphid sequence was derived, not from the targeted Pyrsonympha, but from a different unsequenced gut symbiont. Since the gut fauna is well characterized and known to contain several parabasalans (Kirby, 1934), (Grosovsky and Margulis, 1982), these would be the expected contaminants from the gut fauna. No identified parabasalan sequences were found in these populations of PCR products. This is consistent with the hypothesis that other gut symbionts were excluded by the isolation procedure. The fact that the study was able to detect both an algal and a bacterial contaminant demonstrates that the study is able to detect the presence of containing sequences other than that of the putative pyrsonymphid. Although it is possible that Pyrsonympha was selected against in the amplifications, and therefore the putative pyrsonymphid sequence belongs to a different gut organism, this seems to be unlikely. The primers that were used amplified organisms ranging from bacteria to sub-crown eukaryotes and are predicted to amplify both parabasalans and diplomonads. The possibility that the putative pyrsonymphid sequence belongs to an unidentified and uncharacterized eukaryotic gut symbiont has not been eliminated. However, the well characterized nature of the Reticulitermes gut fauna makes this possibility seem unlikely. The most parsimonious explanation for the assembled data is that I have successfully amplified and sequenced a portion of the small subunit ribosomal R N A gene from Pyrsonympha.  43  44  Figure 5 : Isolation of Pyrsonympha sp. from ft hesperus hindgut Pyrsonympha sp. was identifiable by morphology and organisms assumed to be Trichonympha agilis (A) were excluded.  T e r m i t e gut  1 2  3  1 2  4  3  4  Figure 6 : PCR products from replicates 1 and 2, Chapter Two A= Replicate 1, FJ= Replicate 2 Lane 1: Cercomonas DNA, Lane 2 : 1 Kb ladder size standards, Lane 3: DNA extracted from 50-75 Pyrsonympha and Dinenympha individuals, Lane 4: No-DNA control.  45  AGGGCACCACCAGATGTGGAGCATGCGGCTCAATTCGACTCAA CACGGGGAACTTATCAG  60  G T T C G G A C A C G G T G A G G A T T G A C A G A T T G A T A G C T C T T T C T T G A T T C C G T GGGTGGTGGT  120  GCATGGTCGTTC TTAGTTGGTGGGGT G G C T T G T C T G C T T A A T T G G G A T A A C G AACGAGAC  180  TTCGATTTTTAACTAGGGGGGC GC C AC GGCTC GAAAGTGTCGTGGTG TTTTTACCCCTTC  240  TTAAAGAGACTACCGTTGAGAAACGGGGGAAGTGAGGCAATAACAGGTCCGTTATGCCCT  300  TAGATAATC TGAAC C GC ACGC GTAC TACTC TAGC C GC CTC AATGAGTTTGTGGTGGC CC G  360  GAAGGGTTGCCTACTCTTTTTAGTAGCGGCTGTGCTGGGGATCGGTTTTTGTAATTTCTG  420  AACNCGTGAACGCGGAATATCTCGTAAGTGCAGTTCATCAGACTGCGCTGATTACGTCCC  480  TGCCCTTGTACACACCGCCCGTCGCTCCTACCGATGAATGATCAGGTGAAATGCTCGGAT  540  AGAACAGTTGTGGCGGGTCTCTCGGGGTCCGCTGCGATCGCTCCAGAAGAGCAGTAAATC  600  TTATTAT T T AGAGGA A G GA G A A G T C G T A A C A A G G T T T T C  Figure 7: Sequence of putative pyrsonymphid.  Sequence in bold represents regions where both strands are in agreement, whereas plain sequence represents areas of some disagreement between the two strands. The underlined region represents sequence from only one strand.  46  639  Sum  Smallest Probability  A producing  H i g h - s c o r i n g Segment P a i r s :  Neobulgaria  pura  var. f o l i a c e a  P(N) s m a l l . . . 2 l e - 77  rib... 2 4 Chaetomium elatum 18S r i b o s o m a l RNA. Botryosphaeria ribis 18S s m a l l subun... 5 Pseudalllescheria boydii p a r t i a l 18S... 2 Hypocrea lutea DNA f o r 18S r i b o s o m a l . . . 2 P. anserina mRNA f o r 1 6 S - l i k e RNA 2 Bunodosoma granulifera 18S r i b o s o m a l . . . 4 Tuber c f . rapaeodorum gene f o r 18S r . . . 3 Aureobasidium  pullulans  Taxonomy  16S-like  9e2e4e4e5e5e4ele-  77 77 77 76 76 76 75 74  F . F F F F F F An P  B producing  H i g h - s c o r i n g Segment P a i r s : 18S rRNA gene 18S, ITS-1, 5.8... eukaryote c l o n e v a d i n E . . .  1 4 Unidentified 7 Agaricostilbum hyphaenes 18S r i b o s o . . . 2 A.domesticus 18S rRNA gene and i n t e . . . 3 7 Neoechinorhynchus pseudemydis 18S r..-. C.prolifera (84.28) gene f o r 18S r i . . . 1 Entomophthora Entomophaga  C  P(N)  producing  muscae  aulicae  H i g h - s c o r i n g Segment P a i r s :  Komma Dicyema Dicyema Dicyema  R.mariana Devescovina  09 06 06 05 05 05 04  F F  -  F An An P  P(N)  nucleomorph 18S r i b o s o . . . 18S rDNA sp. 18S rRNA gene orientale gene f o r 18S rRNA,... acuticephalum gene f o r 18S r . . . nucleus.gene f o r SSU rRNA,... sp. 1 6 S - l i k e SSU rRNA gene  caudata  Nanochlorum  3ele7e9e6e8e2e-  eucaryotum  1 4 6 7 8 1 1  lele6e5e6e2e3e-  16 15 15 15 15 14 14  Ch P An An An Ch Pr  D producing  P(N)  H i g h - s c o r i n g Segment P a i r s :  B.vulgaris  5.8S ribosomal RNA,  C.microphylla  5.8S  ITS1...  r i b o s o m a l RNA,  I...  s m a l l subunit r . . . DNA f o r 18S r i b . . . Arabiddpsis thaliana 18S rRNA gene. Arabidopsis thaliana genes f o r 5.8S... V a l e n c i a orange rRNA gene Bangia  atropurpurea  Bangia  atropurpurea  2 .8e- •12 2 8e- 12 4 7e- 12 4 7e- 12 1 8e- 11 1 .8e- 11 1 9e-•11  P P P P P P P  Figure 8 : Top seven BLAST hits from the putative pyrsonymphid A= The 444 bp from the 5' end. The animal and plant hits were not in the top 7 but were included to show that there was little rise in p values between animal, plant and fungus. B= 132 bp region from the interior of the gene, C= positions 419-519, D= the 100 bases from the 3' end. The taxonomic classification of each match is given, F=Fungi, An=Animal, P=Plant, Pr=Protozoa, Ch=Chromist  47  , Mc.jannasc Methaoococ L.ccasei ^.subtilis . Giardia m 3  1001 10C  iMa !  100" 1001  93i  d'Plomonads  HSU1751Z HSU17508  100  }  S  itrlchomo icnomona , Monocercom _oronympha — — g. Metadevesc  Parabasalans  Percolozoa  , Psalterio  aflR'a^jf]  toxauda _ BTrypanoso FetaloiTio Dona , cons.Peran i E uglena gr Khwawkinea itamoeoa ntamoeba Tfireafamoe _ _ Qictyost  100  4  — — 17 12  I  }  Euglenozoa  Amoebozoa  P^hysarumD R L h e s C ^ llassic^ , Collozou pnaerozou ollospnae , Urosponaj Haplospor  iori inchinia Haliommati chaunacant _ Plasmodioj WuChlon e^gmita  na.  ator  ZZZ PaSltneiia  ^Jg|rcomon  ercpmon PerkinsusL • Prorocentj  Alveolates  ineneria , Sarcpcysti  qnas  Heterdkonts  hucus gist , icfyocha elagomona  f1roteromon Blastecy ^W^/Rhodophytes  :mi|iani_ ia. avlova af raviui 3  • M M Apusomonas  * , Corallochy , uermocyj  ,Gla  , Microciona Ahemonia.s _ _ _ _ _ Mn ,/nemiopsis , uiomug.rno5 bRyJridium Schizosacc  canthoeco lapnanoec  so/osperm QMnyopno  Animals  } Fung  ogochaet _ _ , Cyanophora rZea mays, lan'toniell Nephroselm , Hartmannel' , Acanthamoe Chilornonas aRnodomona , lionio monas  / Plants  Figure 9 : Parsimony tree, Chapter Two  } Cryptomonads  The putative pyrsonymphid is denoted R.hes GS (R. hesperus Gut Symbiont). Boxed area indicates a monophyietic clade encompassing all sequences to the exclusion of the diplomonads and parabasalans 100% of the time. Sequence input order was randomized .3 x using the jumble command. Branch lengths do not represent evolutionary distance. 48  _ B.subtilis U.casei  100  Figure 10 :Neighbour joining distance tree, Chapter Two All sequences within the boxed area, including the putative pyrsonymphid denoted Ft. hes G S [R. hesperus gut symbiont), form a monophyletic group to the exclusion of the diplomonads and parabasalans 100% of the time. The length of the scale line represents one change per 10 nucleotides. Distance matrices were calculated using a Kimura-2 rate variation, and sequences were jumbled. 49  1 2 3 4 5 67  8 9 10.1 1121314  Figure 11 : Restriction analysis of clones P12, P7, P4, P9  Lanes 1-3 are clone P12 cut with no enzyme, Ban\, and Rsal respectively, Lanes 4-6 are clone P7 cut with no enzyme, Ban\, and Rsal respectively, Lanes 7-9 are clone P4 cut with Ban\, Rsa\, and no enzyme respectively, Lanes 10-12 are clone P9 cut with BanI, Rsal, and no enzyme respectively, Lanes 13 and 14 have Ban\ and Rsal respectively with no DNA to ensure no contamination . Size standards were not included. Therefore only conclusions as to the similarity of the products in the different lanes, rather than their sizes, can be inferred The additional higher Mw bands in Lane 2 are assumed to be experimental artifacts, as they are not present in other digests of P12 (Figure 13)  50  1  2  3 4  5  6  7  3.0 kb 2.0 kb 1.6 kb 1.0 kb  500 bp  Figure 12 : PCR products from replicates 3 and 4, Chapter Two The product in PCR replicate 3 was reamplified in PCR replicate 4. This reamplified product is shown here. Lane 1,2,4,5= Reamplification of the PCR products from the third amplification of Pyrsonympha DNA, diluted 1 in 10, 1 in 100, 1 in 1000, and 1 in 10000. Lane 3 is the 1 kb ladder. Lane 6 is Cercomonas DNA (positive control) and lane 7 is the no-DNA control. 51  3.0 kb  2.0 kb 1.6 kb 1.0 kb 500 bp  Figure 13 : Restriction of clone from PCR replicate 3, Chapter Two Lanes 1, 3, and 4 are the putative pyrsonymphid clone cut with Banl, Rsal, and no enzyme respectively. Lanes 5, 7,and 8 are clone from replicate 3 cut with Banl, Rsal, and no enzyme respectively. Lanes 2 and 6 are the 1 kb ladder size standards.  52  1 2 3 4 5 6 7 8 9 1011 1213141516171819 20  1 205 bp  786 bp 650 bp  Figure 14 : Restriction digestion of clones from PCR replicate 4, Chapter Two This figure shows clones from all three classes. Lanes 3, 8, 20 show the putative pyrsonymphid (class I) cut with Ban I, Rsa I, and not cut respectively. Lanes 1, 4 and 5 show a clone from class II cut with Ban I, Rsa I and not cut respectively. Lanes 10, 12, and 13 are clones from class III cut with Ban I. Lanes 7,16, and 18 are clones from class III cut with Rsa I. Lanes 9,15, and 19 are uncut clones from class III. Lanes 2, 11, and 14 show PGEM cut with Apal, Seal, and BspHto yield fragments sized 1205, 786, and 650 bp. The fourth fragment of 356 bp is not visible. The 1 kb ladder is shown in lanes 6 and 17. 53  A Sequences p r o d u c i n g  H i g h - s c o r i n g Segment  Arthrobacter  Bacterial Pseudomonas  Smallest Sum Probability P(N)  Pairs:  ramosus 16S rRNA gene. sp. p a r t i a l 16S rRNA gen. sp. DNA f o r 16S rRNA, .  Pseudomonas  sp. A177  Pseudomonas  Pseudomonas  16S rRNA . mucidolens 16S rRNA ge. synxantha 16S rRNA gen. fluorescens (ATCC 1352. 16S rRNA gene fluorescens 16S rRNA g.  Pseudomonas  sp.  P.aureofaciens  16S rRNA gene sp. IC038 16S ribosoma. corrugata 16S rRNA gene sp. ACAM213 16S r i b o s o . tolaasii 16S rRNA gene  Pseudomonas Pseudomonas Pseudomonas P.fluorescens  Pseudomonas Pseudomonas  Pseudomonas Pseudomonas  16S  7 .6e-39 9 .3e-39 Oe-38 3e-38 4e-38 4e-38 2.4e-38 3 .le-38 le-38 6e-37 5e-37 9e-37 Oe-37 Oe-37 Oe-37 7.Oe-37  ribosomal.  azotoformans  16S  ribosomal  RNA  B Sequences p r o d u c i n g  H i g h - s c o r i n g Segment  S m a l l e s t Sum Probability P(N)  Pairs:  Pseudomonas  ramosus 16S rRNA gene,. sp. p a r t i a l 16S rRNA gene. fluorescens (ATCC 13525.  Pseudomonas Pseudomonas  sp. 16S r i b o s o m a l RNA sp. A177 16S r i b o s o m a l  Pseudomonas  Pseudomonas  16S rRNA g. taetrolens 16S rRNA gen. tolaasii 16S rRNA gene 16S rRNA gene sp. IC038 16S r i b o s o m a l . sp. ACAM213 16S ribosom. pavonaceae 16S rRNA gen. sp. DNA f o r 16S rRNA, p.  Pseudomonas  pseudoalcaligenes  Pseudomonas  agarici  Pseudomonas  stutzeri  Arthrobacter  Bacterial  Pseudomonas Pseudomonas P.aureofaciens Pseudomonas  Pseudomonas Pseudomonas  .  azotoformans  DNA  f.  16S rRNA gene s t r a i n 19smn4 .  l.Oe- 142 2 . l e - 134 4.0e- 134 8.2e- 132 9 .Oe- 130 ,5e- 129 ,8e- 129 ,0e- 128 ,5e- 128 ,8e- 128 ,8e- 128 ,3e- 126 ,9e- 126 ,8e- 125 2 .8e- 125 2 .9e- 125  Figure 15 : BLAST results from sequence of clone L20 A is the result from the first 121 bases of the sequence being sent to GenBank B is the result from the entire sequence being sent to GenBank 54  Sum  Sequences p r o d u c i n g 1 6 S - l i k e rRNA Hibberdia O.danica  carinii...  1 6 S - l i k e r i b o s o m a l ...  papillosa  spinosa  Heterothrix  Botrydiopsis  16S  ribosomal  ...  1 6 S - l i k e ribosomal R... debilis n u c l e a r 18S rib.-. . intercedens  nuclear  18...  18S s m a l l subunit nuc... Botrydium stoloniferum n u c l e a r 18S ... Giraudyopsis stellifera small-subun... Mallomonas akrokomos 18S s m a l l subu... Synura  Taxonomy  . P(N)  palustris small-subun... s m a l l subunit rRNA gene.  Mallomonas Synura  [Pneumocystis  magna  Chrysonephele  H i g h - s c o r i n g Segment P a i r s :  Smallest Probability  uvella  .  7 5 5 5 5 5 5 5 5 5 5 5  7e6e7e7e7e7e7e7e7e7e7e8e-  42 41 41 41 41 41 41 41 41 41 41 41  F. Al/S/Cr Al/S/Cr Al/S/Cr Al/S/Cr/Sy Al/S/Cr/Sy Al/S/ Al/S/ Al/S/Cr/Sy Al/S/ Al/S Al/S/Cr/Sy  1 2e- 24 1 2e- 24 1 2e- 24 . 6 8e- 24 1 9e- •21 4 6e- 21 4 6e-•21  Al/S/Cr Al/S/Cr/Sy Al/S/Cr/Sy Al/S/Cr/Sy F F F  B Sequences p r o d u c i n g O.danica  Mallomonas  c  H i g h - s c o r i n g Segment P a i r s :  s m a l l subunit rRNA gene. papillosa  16S  ribosomal  . . .  18S s m a l l subuni... 18S s m a l l subuni... T.reesei 18S rRNA gene, 297bp. Paecilomyces tenuipes DNA f o r 18S rRNA Y e a s t - l i k e symbiont o f a p h i d (Hamil... Mallomonas  rasilis  Mallomonas  caudata  Sequences p r o d u c i n g Mallomonas  H i g h - s c o r i n g Segment P a i r s :  papillosa  16S  ribosomal  ...  18S s m a l l subuni... Mallomonas annulata 18S s m a l l subun... Mallomonas akrokomos 18S s m a l l subu... Mallomonas caudata 18S s m a l l subuni... Synura uvella 18S s m a l l subunit nuc... Synura petersenii IBS small subunit... Synura glabra 18S s m a l l subunit nuc... Chrysonephele palustris small-subun... Mallomonas  Hibberdia Mallomonas  rasilis  magna adamas  1 6 S - l i k e r i b o s o m a l ...  18S s m a l l  subunit...  P(N)  P(N) 1 .5e- •141 1 .6e- •141 1 .5e- •137 4 .9e- •136 7 .5e- •136 2 .7e- •135 2 .7e- •135 2 .8e- •135 1 .4e- -134 4 .4e- •134 8 .5e- -134  Al/S/Cr/Sy -Al/S/Cr/Sy Al/S/Cr/Sy Al/S/Cr/Sy Al/S/Cr/Sy Al/S/Cr/Sy Al/S/Cr/Sy Al/S/Cr/Sy Al/S/Cr Al/S/Cr Al/S/Cr/Sy  Figure 16 : BLAST results from sequence of clone U39 A is the result from the first 123 base pairs, B is from the middle 111 base pairs and G is from a query using the entire sequence. The taxonomic classification of the organism from which the sequence is derived is also shown. AI=Algae, S=Stramenopiles, Cr=Chrysophyceae, Sy= Synurales and F=Fungi. 55  Organism  putative pyrsonymphid  R.F.G.S 1  R.F.G.S. 2  Fragment size Orientation 1 B 1286 R 2029 1097 1510 687 165 436 198 B 1286 R 1991 1269 1391 1097 270 B 1286 1259 1097  R 1987 1629 26  Orientation 2 B 1286 R 1097 887 236 198 B 1286 R 1269 1097 B 1286 1259 1097  Table 2 : Predicted fragment size from likely gut symbionts. B= Predicted sizes of the fragments if the plasmid were cut with BanI, R= Predicted sizes of the fragments if the plasmid were cut with Rsal.  56  R  2185 1354 165  2066 1316 270 2304 1312 26  class  1  II  III  clones  Fragment size from Banl  Fragment size from Rsal  U38, L7, U11, P12,  1250, 1100, 825, 425  2000, <1600  U39, L4, U13  1250, 1100, 825, 425  L10, L14, L20, L9, L22, L13  1250, 1100, 700, 550  2000, 1600  2000, <1600  Table 3 : Three classes of clones from PCR replicate 4 Included is the fragment size when cut with Banl and Rsal as well as the clones that fall into each class. <1600 denotes the presence of a fragment slightly smaller than 1600bp.  57  Chapter III: Phylogenetic placement of hypermastigotes Introduction As with the oxymonads, the hypermastigotes are sexual (Cleveland, 1949), (Cleveland, 1951), (Cleveland, 1956) and presumed to be ancient based on morphology (Brugerolle and Taylor, 1977). Like all parabasalans, the hypermastigotes are amitochondrial protists that contain parabasal bodies, dictyosomes combined with bundles of cytoskeletal elements. The hypermastigotes are distinguished from the other major group of parabasalans, the trichomonads, by their number of flagella (hundreds to thousands rather than three to five), multiple parabasal bodies, and a distinctive pattern of flagellar organization (all anterior or in anteriorly-meeting rows) (Lee, Hutner and Bovee, 1985). The large hypermastigote Trichonympha is a major hindgut constituent of the termite (Kirby, 1932), and has been shown to be partly responsible for the subterranean termite's ability to digest wood (Grosovsky and Margulis, 1982). The parabasalans are the third eukaryotic branch to diverge in most ssu rRNA trees (Gunderson et al, 1995), (Cavalier-Smith and Chao, 1996). Recent evidence based on tubulin (Keeling and Doolittle, 1996) and HSP 70 sequence data (Germot, Philippe and Le Guyader, 1997), however, suggests that the microsporidia, which normally form the second branch (Cavalier-Smith, 1993), (Cavalier-Smith and Chao, 1996) may be highly derived fungi leaving the parabasalans as the second branch to have diverged.  Trichonympha morphology Unless noted, the information in this morphological description is based on Kirby's 1932 paper on Trichonympha (Kirby, 1932). The genus Trichonympha is subdivided into two groups. T. agilis and T. minor, which compose the agilis group, are small, with lengths ranging from 54 urn to 115 |im (Kirby, 1932). The magna  58  group of T. magna, T. turkestanica, T. campanula and T. collaris, are much longer, ranging from 150 to 360 urn (Kirby, 1932). Trichonympha tends to be radially symmetrical. From the tapered anterior end, the cells widen to the middle section and may taper again to the posterior. The body is divided into three regions descending from the anterior (Fig 17A). The rostrum, a structure at the anterior tip, has its topmost portion covered by a non-flagellated cap. Below this is a flagellated region and the flagellar organizing center, the blepharoplast or parabasal lamella. This complex of cytoskeletal elements contains the primary basal bodies and organizes the flagella that run along the body. A cleft along the surface separates the rostrum from the rest of the cell surface. However, a rostral tube descends into the flagellated mid-region of the cell. This tube connects to the nucleus and is formed of fused basal bodies. The attractophore, made of striated fibers, runs down the middle and acts as an organizing center in mitosis (Hollande and Carruette-Valentin, 1971). The presence of the parabasal lamella and attractophore demonstrate the close affinity of Trichonympha to other hypermastigotes. Two types of flagella are seen in Trichonympha. The short anteriorlypointing flagella originate from the rostrum. However, the much larger flagellated region of the body has elongated flagella that trail posteriorly and tend to be longer the further they are down the body. These originate from the parabasal lamella whose two hemispherical domains extend down along the body in ribbon-like strips. Hypermastigotes have many flagella and Trichonympha is no exception. Depending on the species there may be several hundred to ten thousand flagella, each with a separate basal body attached to the parabasal lamella. In some species there is a region of very short flagella in the middle of this region. Although larger species tend to have larger nuclei, nuclear size is not tightly correlated to body size. In addition to regular chromatin, some species have a clear  59  region in their nucleus, interpreted to be a nucleolus. The nucleus is positioned near the boundary between the flagellated mid-region and the posterior nonflagellated region. The posterior end of the cell is the site of the feeding and sexual apparati. This region varies in size and also contains food vacuoles. The cell is also organized in layers from the cell surface into the interior (Fig 17B). Flagella protrude from between ridges on the cell surface. Beneath this surface lies the ectoplasm. This outer cytoplasmic region is subdivided into three layers, outer, middle and inner whose relative thickness varies between species and regions of the body. The innermost ectoplasmic layer contains the ribbon-like bands of the parabasal lamella from which the individual flagella originate. The endoplasm contains the visible organelles including the parabasal bodies which form a basket-like structure around the nucleus. Parabasal filaments connect the bodies to the parabasal lamella. A variety of granules are also present in the endoplasm which may represent storage products, smaller organelles hydrogenosomes and endozoic bacteria. The parabasal bodies and hydrogenosomes are shared by all parabasalans. There are no mitochondria in Trichonympha or any other parabasalan. This lack has been used, among other characteristics to infer that they are an ancient lineage.  Sexual cycle of Trichonympha For inferences about the ancestral sexual cycle to be made from that of the hypermastigotes, the sexual cycle must share a common origin with that of the rest of eukaryotes. It has been assumed in this thesis that all sexual cycles evolved from one evolutionary event. This is based, among other evidence, on the common elements throughout the sexual cycles of eukaryotes. The following description will  60  show that, although variation exists, the sexual cycle of Trichonympha contains the same essential components as all other sexual cycles. The sexual cycle of Trichonympha has the three stages common to all eukaryotic sexual cycles, gametogenesis, fertilization, and meiosis (Fig 18), reproduced from (Cleveland, 1956). Gametogenesis in species inhabiting termites and those inhabiting the wood eating cockroach Cryptocercus punctulatus differ and will be treated separately. In Trichonympha that inhabit termites, gametogenesis is similar to mitosis, with duplication of the cytoskeletal system, and subsequent cytoplasmic division. However, gametogenesis yields two cells of unequal size, which behave differently in the subsequent gamete fusion. In Trichonympha that inhabit Cryptocercus, both the cytoskeletal system and chromosomes duplicate, the latter forming two sets that differ in their degree of chromatin condensation. A cyst forms that protects the cells when they are ejected from the host during the molting process. After passing to a new host, excystation occurs with cytoplasmic division yielding two gametes. Although not true egg and sperm, Cleveland described the anisogamous gametes of Trichonympha as such (Cleveland, 1949). The larger egg possesses a fertilization ring at the cell's posterior end from which the cone of fertilization extends (Fig 18). The sperm is smaller with several heavily stainable granules in its dense cytoplasm. In the termite gut gametes swim side by side, fusing at their anterior tips. In Trichonympha from Cryptocercus the sperm attaches to the cone of fertilization and is pulled inside the egg. It loses part of its cytoplasm and all of its organelles. In Trichonympha from both hosts, the sperm nucleus migrates towards the egg nucleus. These fuse after the formation of a thick common membrane around the two nuclei.  61  Persistence of the zygote varies from species to species. After some period of time, the chromosomes duplicate, and homologues pair. This is followed by what appears to be standard two-step meiosis to yield four asexual daughter cells.  Zootermopsis gut fauna Trichonympha can be found locally in the guts of Zootermopsis angusticollis. These large rotten wood termites are endemic to the lower mainland and found in logs on the beaches near the University of British Columbia. Z . angusticollis has a well characterized gut fauna which includes three species of Trichonympha (Kirby, 1932). These are distinguished by size and shape as well as several internal characteristics. The gut is also home to three trichomonads, Tricercomitus termopsidis and Hexamastix termopsidis and Trichomonas termopsidis Sharing the gut with the parabasalans is the oxymonad, Streblomastix strix. Gregarines (Apicomplexans) are reported to inhabit some colonies but were not seen in the course of my study.  Why hypermastigotes are candidates as the deepest sexual lineage. Due to the strong morphological similarities of the parabasalan classes, it is very likely that the hypermastigotes and the trichomonads are closely related. Shared characteristics include the presence of parabasal bodies, hydrogenosomes and a stationary axostyle, as well as the absence of mitochondria. The proposed evolution of the parabasalans has been illustrated in a 1977 symposium review (Brugerolle and Taylor, 1977), redrawn in Figure 19 as a phylogenetic tree. The hypermastigotes are the most complex group and are proposed to be derived from within the trichomonad branch.  62  As convergence is always a possibility when classifying based on morphological characteristics, it is desirable to confirm that the morphological similarities within this clade are due to common descent. This can be addressed by molecular phylogenetics.  63  Materials and Methods F i n d i n g , identifying and keeping termites The beaches of Spanish Banks, Vancouver B.C., were searched for Zootermopsis angusticollis. Logs in the middle stages of decomposition and with termite holes of 3 to 5 mm in diameter were deemed likely to contain a colony. Termite pellets that were firm, without being too dry, indicated recent or current inhabitation of the log. When possible the same colony was used to collect samples, however storms would occasionally rearrange the logs on the beach and so new colonies had to be found in several instances. Z. angusticollis individuals are large, up to 1 cm in length and range from a cream to dark brown colour depending on season and time since feeding. Since they are the only termites to inhabit rotten wood in this geographic region (Ruppel, 1978) there was no need for stringent morphological identification. Along with pieces of wood for habitat and food, one to ten individuals were collected per sampling run. The specimens were kept in a 500 ml plastic bottle with a plastic bag over the opening. A n elastic band kept the cover tight enough to prevent escape but allowed for airflow and moisture exchange.  Dissection and identification of Trichonympha Figure 20 illustrates the washing protocol used to isolate Trichonympha individuals. The first step was to empty the gut fauna from a decapitated Zootermopsis worker into 500 ul of modified Trager's medium (Buhse, Stamler and Smith, 1975). In step two an aliquot of this diluted gut fluid was transferred into 500 ul of fresh medium and was viewed with an Olympus dissection microscope at maximum magnification. These two steps provided dilution of the gut fauna. During the third round of dilution, cells could be separated by size, and the large Trichonympha cells were drawn into a blown glass micropipette and away from the  64  smaller protists. In step four, Trichonympha of different sizes and length to width ratios were clearly distinguishable and so the broader Trichonympha cells (tentatively identified as Trichonympha cf. collaris) were chosen. This process was repeated once more in step five before pooling the individuals into a 1.5 ml microfuge tube. This final round of dilution served to confirm the identification of the Trichonympha species and to further reduce the number of errant protists transferred into the solution. Measurements of cells selected under the dissecting scope were made under 400x magnification (Fig 21). The 500 | i l pools of media were replaced after every 50 organisms were selected. In total 350 cells were collected.  D N A extraction  The DNA extraction protocol was identical to that used for the oxymonad project. See Chapter Two, Materials and Methods, DNA extraction.  PCR  Primer synthesis and concentration, total reaction volume and dNTP concentration were as described in Chapter Two, Materials and Methods, PCR. Eukaryotic specific primers were used (Gunderson et al, 1995). Taq reaction buffer (Gibco BRL) was used at lx concentration and M g  2 +  (Gibco BRL) was added to a final  concentration of ImM. Fifty nanograms of DNA from Paramecium tetraurelia, provided by Hong Zhang, were used in positive control reactions. Sterile water from the Milli-RO 60 Reverse Osmosis system (ZFNC 115 60) / Milli-Q Plus Water Systems (ZD40 115 95) was used as a no-DNA control. DNA was extracted from a 400 u.1 aliquot of the final Trichonympha isolation wash (step five, Fig 20) and used as a  65  second negative control. ART 200 aerosol tips (Continental Laboratory Products) were used for all micropipetting. Each reaction was set up on ice in a 500 ul PCR tube and performed using a Perkin Elmer Cetus D N A Thermal Cycler 480. Reactions began with an initial cycle of 94 °C for 1 minute, followed by 1 minute of 37°C and 3 minutes of 72 °C. This cycle was repeated 43 more times with the initial heating step at 94 °C lasting for 10 seconds. A final cycle of 94 °C for 10 seconds, 1 minute at 37°C and 4 minutes at 72 °C was run to promote the complete synthesis of PCR products.  PCR product extraction and purification PCR product extraction and purification were carried out as described in Materials and Methods of Chapter Two, PCR product extraction and purification.  Cloning All vectors, protocols and strains were as described in Chapter Two, Materials and Methods, Cloning section.  Clone selection Clone selection, screening and plasmid preparation were done as described in Chapter Two. Plasmids were linearized by digestion with Apal (Pharmacia Biotech), and visualized by electrophoresis on a 0.8% agarose gel. Both 1 kb ladder (Gibco BRL) and Lambda Hind III D N A (Gibco BRL) were used as size standards.  Sequencing Plasmid purification, sequencing, and determination of regions of reliability were done as described in Chapter Two. Initial sequence was obtained from the M13 forward and reverse sequencing primer sites in the vector. Internal sequencing  66  primers, shown in Table 4, were designed to fit a region of reliable sequence from the previous run. These internal primers were 16 or 17 bp long and ended with a GC to ensure strong 3' base pairing. The final sequence of Trichonympha cf. collaris has been placed in GenBank with accession number AP023622.  Phylogenetics Blast analysis was done using the default settings at the NCBI Basic BLAST database search page (http://www.ncbi.nlm.nih.gov/BLAST/). Clustal W (Thompson, Higgins and Gibson, 1994) was used initially to align 18 ssu rRNA sequences which were then adjusted by eye. All species used along with their accession numbers can be seen in Table 5. PHYLIP version 3.57c (Felsenstein, 1995) was used for all phylogenetic analysis. The alignment was bootstrapped 100 times, and rooted with the bacterial sequence as the outgroup. Any parameter settings not set at default values will be stated on the figures. Trees were visualized using the TreeView™ program (Page, 1996).  fluorescence in-situ hybridization FISH In order to design a probe specific for the T. cf. collaris sequence, it was aligned with the ssu rRNA sequences of Trichomonas tenax, Coronympha octonaria, and R. flavipes gut symbionts 1 and 2. I located variable regions in the alignment that were bracketed by conserved regions. This verified that apparent variability was not due to mis-alignment. A solution of 0.5% gelatin and 0.03% chrome-alum was prepared by dissolving the compounds at 65 °C. Slides were dipped in the solution for 20 seconds at 40 °C, allowed to air dry and kept at 4°C until use. Gut fauna from a single Zootermopsis worker was emptied into a fixation solution of 10%  67  Formaldehyde in PBS diluted 1/9 in modified Trager's medium (Buhse, Stamler and Smith, 1975). A sample of the fixed gut fauna was then dried onto a pre-treated slide, fluorescence in-situ hybridization (FISH) was carried out according to the protocol of Giovannoni et al (Giovannoni et al, 1988). All cells were probed with 30 ul of either the universal probe (5'-GA/TAATACCGCGGCG/TGCTG-3') or a probe designed to be specific to the T. cf. collaris sequence (5 '-T A G A G A T A A A T C C T T A-3') at a concentration of 1.7 ng/ml. All probes were labeled with a fluorescein molecule at the 5' end. Thirty microlitres of the solution in which the probes were dissolved was used as a no-probe control. Airtight Tupperware containers, freshly covered with tin foil to block out light, held the probed slides during the washing step. This took place in a water bath shaker at 37°C. Slides were air dried at 37°C in the dark. Images were obtained on a Zeiss PhotoMicroscope with integrated automatic 35 mm camera using T M A X 400 film. A 520 nm barrier filter was used to maximize detection of the emission wavelength (495 nm) of fluorescein. Photographs were printed using standard darkroom printing techniques in the U.B.C. Biosciences Electron Microscopy Facility. All fluorescence micrographs, were taken using 4 seconds shutter time, and exposed for 1.5 seconds. The no probe control was exposed for 1.9 seconds to emphasize that no background fluorescence was seen in anterior portions of Trichonympha.  68  Results Identification of organisms A pure sample of Trichonympha was needed to amplify the ssu r D N A gene. This sample was obtained by identification and dissection of the termite host and microscopic identification of the protists. Kirby described three Trichonympha species in Z . angusticollis :T. campanula, T. sphaerica, and T. collaris (Kirby, 1932). However, he reported substantial overlap in sizes, between T. collaris (247x114 urn) and T. campanula (217x85 urn). I identified the three different types of Trichonympha based on cell size and length to width ratios, and tentatively assigned the largest and broadest cells to be Trichonympha collaris. Based on four measured individuals, the largest type measured 252 um long (with a range of 240-270 urn) and averaged 217 urn wide (range 210-230 urn). A second type, tentatively identified as T. campanula, measured 256 urn long (with lengths ranging from 210-300 urn) and 146 urn wide (the widths ranged from 130-170 urn). This and the third type identified as T. sphaerica (126 u m x 110 urn) were excluded by the washing procedure. Ranges in length and width for the tentative T. sphaerica were 105-170 urn and 75-150 u.m and are based on five measured individuals, as was also the case for the tentative T. campanula. The two largest cell types of Trichonympha were not reliably distinguishable under the dissecting microscope used for collecting the cells and so the Trichonympha species that I isolated is only tentatively assigned to Trichonympha collaris (Trichonympha cf. collaris). The washing steps allowed both for dilution of unwanted gut protozoa away from the targeted Trichonympha cells and confirmation of the morphological identification of the Trichonympha species. Because of the extent of the washing procedure, it is likely that primarily Trichonympha cf. collaris D N A was extracted.  69  PCR PCR amplification of the extracted Trichonympha cf. collaris D N A yielded a single 1.5 kb band (Fig 22). The Paramecium tetraurelia D N A gave a 1.7 kb band, the predicted size, and the no-DNA control showed no product. A faint band was seen in the gut fluid sample (approximately 1.5 kb) compared with the approximately 1.55 kb band in the Trichonympha  sample.  The strength of the single band in the Trichonympha sample and the lack of bands in the no-DNA control, along with the faintness and difference in size of the gut fluid band indicated that the amplification of the Trichonympha cf. collaris D N A was likely to have been successful.  DNA preparation and Cloning Due to the presence of a smear of D N A around the Trichonympha  PCR  product, and presence of the gut fluid band, it is possible that additional organisms may have been present in the sample of Trichonympha DNA. Therefore it was necessary to purify the PCR fragment and clone it. The difference in size between the band in lanes 1 and 3 in Figure 22 allowed for resolution of the two bands. This reduced the likelihood of contamination in the cloning process. The 1.55 kb PCR product was purified (Gene Clean II) and cloned into E. coli DH5oc. Plasmid D N A prepared from clones that possessed the disrupted vector was further screened by restriction analysis. When cut with Apal clone T3 yielded a single band of 4.5 kb, corresponding to the insert size (1.5 kb) plus the vector (3.0 kb). This clone was used for subsequent experiments.  70  Sequencing To obtain the sequence of the insert in clone T3, plasmid D N A was prepared by lithium chloride purification and sequenced by the D N A sequencing facility at UBC. The contiguous sequence from both strands of clone T3 sequence is 1518 bp long. Because PCR initiates from the 3' end of each primer, the T3 sequence contains all but approximately 25 nucleotides at each of the 5' and 3' ends of the small subunit ribosomal rRNA gene (Fig 23). BLAST analysis of two portions of the T3 sequence against the GenBank database showed the strongest sequence similarity to the ssu rRNA sequence of the R. flavipes gut symbiont 2 (GenBank accession number HSU17512) (Gunderson et al, 1995) with p values of 2.3x I O  - 1 0 7  and 1.8xl0' respectively (Fig 24). The 98  Trichonympha cf. collaris sequence is 92.3% identical to R. flavipes gut symbiont 2 over 1525 aligned positions (Fig 25). Among the hypermastigotes in the hindgut fauna of R. flavipes is Trichonympha agilis (Grosovsky and Margulis, 1982). Based on the low p value for the BLAST matches and the percent identity to the T. cf. collaris sequence, it seems reasonable to assume that the R.flavipesgut symbiont 2 sequence belongs to a Trichonympha species. This will be further discussed in Chapter Four. It should be noted that late in the course of my thesis work Ohkuma et al independently identified a Trichonympha ssu rRNA sequence using whole gut PCR from Reticulitermes speratus and fluorescence in-situ hybridization (Ohkuma et al, 1997). This sequence was 93.6% identical to the R. flavipes gut symbiont 2 and 92.4% identical to the T. cf. collaris sequence.  Phylogeny A phylogenetic analysis was done to quantify the relationship of the T. cf. collaris sequence to R. flavipes gut symbiont 2, as well as the relationship of the  71  hypermastigotes to the rest of the parabasalans. The Trichonympha cf. collaris rRNA sequence was aligned with that of 14 other parabasalans, two diplomonads and an archaebacterium. As can be seen in the alignment, available at http://www.zoology.ubc.ca/~redfield/hypermas.html, the sequences were easily alignable over their length, with only a few regions of high variability. This alignment was bootstrapped and analyzed using both neighbour joining (Fig 26) and parsimony tree methods (Fig 27). The two Trichonympha sequences formed a monophyletic clade with the R. flavipes gut symbiont 2, 100% of the time using both methods. High bootstrap values also excluded the hypermastigotes from the monophyletic clade formed by the trichomonads. Implications of the placement of the hypermastigotes are discussed in the final chapter.  FISH Although the evidence for the T. cf. collaris sequence belonging to a Trichonympha is strong, there still exists a small possibility that the sequence belongs to another gut organism. To eliminate this possibility, FISH studies (DeLong, Wickham and Pace, 1989) were performed on the gut fauna of Z. angusticollis. The hindgut contents of a Zootermopsis angusticollis were fixed and probed with a universal positive control probe (Giovannoni et al, 1988), and a probe specific to the variable region in the T. cf. collaris sequence. As seen in Figure 28, the universal probe bound strongly to almost the entire length of the Trichonympha cells. Figure 29 shows fluorescence in the posterior end of the Trichonympha cells with a no-probe control. Since wood autofluoresces and the fluorescence in this figure is restricted to the region of the cell containing wood particles, this signal is presumed to be due to autofluorescence.  72  The probe derived from the T. cf. collaris  sequence bound strongly to approximately four fifths of the length of the Trichonympha cell (Fig 30). However, some anomalies existed in the FISH data that need to be addressed. Figure 30 shows Trichonympha cells strongly fluorescing after being incubated with the specific probe and adjacent Streblomastix cells without fluorescence. However, this result is complicated by the fact that Streblomastix cells bound the universal probe much less than did the Trichonympha cells (Fig 31). Additionally the problem of autofluorescing wood in the gut environment and fauna made interpretation of the fluorescence data difficult (Fig 32). These factors detract from the certainty of the FISH data but do not seem to be enough to discount the strength of specific binding to the Trichonympha cells. In summary, from all of the evidence collected, phylogenetic and FISH, it seems very likely that the sequence that I obtained belongs to Trichonympha, and therefore can be confidently used in phylogenetic analysis.  73  74  Figure 17 : Trichonympha cf collaris. This artist's rendition was drawn by Tamara Hartson from textual descriptions and observations of live material.  A  Figure 18 : Sexual cycle of Trichonympha from Cryptocercus reference in text.  entamoeba  | 1  Moriocercomonas  Calonymphidae  Trichomonads  i chomonas  Devescovi ni dael  Trichomitus  Pentatrichomonas  Joeni ds  Figure 19 : Evolution of the Parabasalans Redrawn from Brugerolle and Taylor (1977). This tree assumes morphological simplicity to be the ancestral state. 76  Hypotri chomonas  HexamastiXi  Histomonasfj  Monocercomonads  Trichomopsis  77  Trichonympha cf. collaris was identifiable by morphology and organisms assumed to be T. campanula (A) and T. sphaerica (not shown)were excluded.  Figure 20 : Isolation of T. cf collaris from Z. angusticollis  Termite gut  Figure 21 : Trichonympha sp. viewed using phase contrast microscopy viewed at 40X magnification.  78  1 2 3  4 5  1.7kb 1.5kb  Figure 22 : PCR products from amplification using eukaryotic specific primers Lane ~\-.Trichonympha DNA template, Lane 2: 1 kb ladder size standards Lane 3: Gut fluid control, Lane 4: Paramecium DNA template, Lane 5: No DNA control  79  GAAGCACACTTCGGTCAGAGATGAAGCCATGCAAGTGCTAGTTAAAGTAATAAACTGCGA • '  •  •  •  .  *  60  *  ACAGCTCATTMCACACTCAGGGTCTMTTGATTGTGACATTATTGTTTAATATGGATAG •  •  •  •  •  • •  •  180  TCGGATTAAATCTCGGACTMTACATGCAATTGTTTCATCAAAATCGAAATATGATGGAA •  •  •  •  .  •  •  MGTTTACCCTCACGGGCAMTCATTGGATTGAGTGTCCTATCAGCTATTAAGTAGGGTC •  •  •  •  120  •  • •  240  TTTACCTATTTAGGCTATTACGGGTAACGGGTAAGGGGCGGTTGTCGTCGAACTGCCGGA 300 •  •  '  •  -  •  •  •  GAAGGCGCCTGAGAGATAGCGACTATGTCTACGGAAAGCAGCAGGCGCGAAACTTACCCA  360  CTCGTAMGCACGGATGTGGTTATGACCAGTTCTATMGGGATTTATCTCTAGTAGATAG  420  •  •  •  •  •  •  GAAGATACACAATGACTATCTTGAAATCGAGCAGAGGGCCAGTCTGGTGCCAGCAGCTGC  480  GGTAATTCCAGCTCTGTMGTGTGCTCCCATATTGTTGCAGTTAAAAGACCTGTAATCGG  540  •  •  •  •  •  •  ATTTCAATTGAAMCTAAMTATTCACTGTGMTAAATTAGGACGCTTAAAGTATGGTTG  600  TATGAATACTTTAGCGCAGTATGGAAGATTTTGCTCGCATGGGCAAGATCAAAGAGAGTC  660  ATTGGGGGTATTTCTATTTCATGGCGAGTGGTGAAATATGTTGACCTATGGGAGAGAAAC 720 •  •  •  •  •  •  •  GAAAGCGAAGGCAAATACCCAAAGGGTTTCTATCGATCAAGGGCGAGAGTAGAGGGAGCG 780 •  •  •  •  *  *  AACCGGATCAGAGACCCGGGTAGTCCCTACTGTAAACGATGCCGACAGGGAATTGTCCAC •  •  •  •  •  •  •  900  TTGGTGGACAGMCCTTAGCAAAAATGATAGTTCATGGACTCTGGGGGAACTACGACTGC •  *  •  •  •  840  '  •  •  AAGGCTGAAACTTGAAGGAATTGACGGAAGGGCAGACCAGGGGTGGAGCCTGTGGCCTAA  960  TTTGMTCAACACGGGGAMCTTACCAGMCCAGATGATTCGATGACTTACATCTAAAAG  1020  «  •  •  •  •  •  AGCTTTMGGATGGGTTTGATGGTGGTGCGTGGTCGTTGGTGGTCCGTGGGTTGACCTGT •  •  •  *i  •  1140  CACGAGTCGATTCAGTTAACGAGCGAGACTACCACCAAATATTAGCGATCCTTAGATTAT •  •  •  •  •  •  1200  MGGAMGCTTCTATTTGGGACTACCTGTTTTCAAGCGGGAGGAAGAGGGTAGCAATAAC •  •  •  .  •  *  1080  , •  •  •  AGGTCCGTGGTGTCCTTTAGATGCTCTGGGCTGCACGCGTGCTACAATGTGAATCACAAA  1260  GGAAAMTAAATCGAMGMTGCATCACTTCTCCAAAGATTCACGTAGCTGGGATTGACA  1320  TTTGGMTTATTGTCATGMCCAGGMTCCCTTGTAMTGTATGTGAACATCGMCGTTG  1380  •  •  •  •  •  •  AATACGTCCCTGCCCTTTGTACACACCGCCCGTCGCTCATACCGATTGGACGTTTTGGTG •  •  •  •  •  •  MAATACTGGACTTTMTTTAMGGMGGTATTTAMTCAGCTCGTCTAGAGGMGGAG^ •  AGTCGTAACAAGGTAACG  •  •  •  Figure 23 -.Trichonympha cf. collaris ssu r R N A gene sequence 25 bp removed at the 5'and 3'end. 80  •  •  •  1440 •  1500 1  5  1  8  Sequences p r o d u c i n g  H i g h - s c o r i n g Segment P a i r s :  U n i d e n t i f i e d trichomonad 2 1 6 S - l i k e . Monocercomonas sp. ATCC 50210 1 6 S - l i . Pseudotrichomonas keilini 16S-like r . Ditrichomonas honigbergi 1 6 S - l i k e rR. Coronympha octonaria 1 6 S - l i k e rRNA g. Trichomonas tenax 1 6 S - l i k e r i b o s o m a l . Tritrichomonas foetus 1 6 S - l i k e rRNA . Tritrichomonas foetus IBS, 5.8S, and. U n i d e n t i f i e d trichomonad 1 1 6 S - l i k e . Trichomonas vaginalis 1 6 S - l i k e rRNA . Dientamoeba fragilis 1 6 S - l i k e rRNA, . Metadevescovina polyspira 16S-like r . Trichomonas tenax gene f o r SrRNA Devescovina sp. 1 6 S - l i k e SSU rRNA gene Unknown trichomonad NJ1 1 6 S - l i k e SSU.. P.scroa 18S rRNA gene  Smallest Sum Probability P(N)  6e- 107 Oe- 71 4e- 66 2e- 66 9e- 64 4e- 63 le- 61 8.8e1. 8e3 .0e4. 9e4. 3e-  60 57 57 57 56 1.2e- 54 6.7e- 50 1.8e- 48 9.4e- 48  B Sequences p r o d u c i n g  H i g h - s c o r i n g Segment P a i r s :  U n i d e n t i f i e d trichomonad 2 1 6 S - l i k e .. Ditrichomonas honigbergi 1 6 S - l i k e rR.. Monocercomonas sp. ATCC 50210 1 6 S - l i . . Pseudotrichomonas keilini 16S-like r . . Trichomonas tenax gene f o r SrRNA Trichomonas tenax 1 6 S - l i k e r i b o s o m a l . . P.scroa 18S rRNA gene Trichomonas vaginalis 1 6 S - l i k e rRNA .. Calonympha sp. 1 6 S - l i k e SSU rRNA gene Coronympha octonaria 1 6 S - l i k e rRNA g.. Tritrichomonas foetus 1 6 S - l i k e rRNA .. T.trypanoides (Rl) 18S rRNA gene. Tritrichomonas foetus 18S, 5.8S, and.. M.extranea 18S rRNA gene  Smallest Sum Probability P(N) 1.8e-98 1.3e-66 6.6e-65 8.6e-55 9.8e-53 6.3e-52 2 .le-50 8.1e-49 4.3e-47 1.7e-42 1.6e-40 6.0e-40 7.2e-40 1.2e-35  Figure 24 : BLAST results IromTrlchonympha cf. collaris sequence. A= Results from a query derived from the 5' end of the gene. B= Results from a query derived from the 3' end of the gene.  81  U. I. T. _2 T._cf._col  -  AACCTGGTTGATCCTGCCAGGGAAGCACACTTAGGTCAGGGACTAAGCCATGCAAGTGCT -GAAGCACACTTCGGTCAGAGATGAAGCCATGCAAGTGCT  *********** ****** ** U.I.T._2 ' T ._cf ._col U . I. T . _2 T. _ c f . _ c o l  ****************  AGTTAAAGTAATGAAACTGCGAACAGCTCATTAACACACTCAGAATCTAATTGATTGCGA AGTTAAAGTAAT-AAACTGCGAACAGCTCATTAACACACTCAGGGTCTAATTGATTGTGA ************ ****************************** ************ ** TTGTAATATCTTATATGGATAGTCGGAGTAAATCTCGGACTAATACATGCAATTGTTTCA CATTATTGTTTAATATGGATAGTCGGATTAAATCTCGGACTAATACATGCAATTGTTTCA  ** * * * * * * * * * * * * * * * * * *  ********************************  U. I. T . _2 T._cf._col  TCAGAATCGAAATATGATGGAAAAGGTTGACCCTTAGGGGCAAATCATTGGATTGAGTGT TCAAAATCGAAATATGATGGAAAAG-TTTACCCTCACGGGCAAATCATTGGATTGAGTGT *** ********************* ** ***** * ***********************  U.I.T._2 T ._cf ._col  CCTATCAGCTATTAAGTAGGGTCTTTACCTATTTAGGCTATTACGGGTAACGGG CCTATCAGCTATTAAGTAGGGTCTTTACCTATTTAGGCTATTACGGGTAACGGGTAAGGG ******************************************************  U. I. T . _2 T ._cf ._col U. I. T . _2 T . _ C f . _co 1 U. I. T . _2 T ._cf ._col  -CGGTTG-CGTCGGACTGCCGGAGAAGGCGCCTGAGAGATAGCAGCTATGTCTACGGACA GCGGTTGTCGTCGAACTGCCGGAGAAGGCGCCTGAGAGATAGCGACTATGTCTACGGAAA ****** ***** ***************************** ************* * GCAGCAGGCGCGAAACTTACCCACTCGTAAAGCACGGAGGTGGTTATGACCAGTTCTATA GCAGCAGGCGCGAAACTTACCCACTCGTAAAGCACGGATGTGGTTATGACCAGTTCTATA  **************************************  *********************  TGTGAGTTAATCTCAGGTAGATAGGAAGATACCAAAAGACTATCTTGAAATCGAGCAGAG AGGGATTTA-TCTCTAGTAGATAGGAAGATACACAATGACTATCTTGAAATCGAGCAGAG * ** *** **** **************** ** ***********************  U.I.T._2 T._cf._col  GGCCAGTCTGGTGCCAGCAGCTGCGGTAATTCCAGCTCTGTAAGTGTGCTCCCATATTGT GGCCAGTCTGGTGCCAGCAGCTGCGGTAATTCCAGCTCTGTAAGTGTGCTCCCATATTGT ************************************************************  U. I. T . _2 T._C f . _ C O 1  TGCAGTTAAAAAGCCCGTAGTCGGATTTCAATTGA CTAA-TTATTCACTGTGAATAA TGCAGTTAAAAGACCTGTAATCGGATTTCAATTGAAAACTAAAATATTCACTGTGAATAA ***********. ** *** *************** **** ****************.  U. I. T. _2 T ._cf ._col  ATTAGGACGCTTAAAGTATGGTTGCATGAATGACTT-AGCGCAGTATGAAAGATTTTGCNC ATTAGGACGCTTAAAGTATGGTTGTATGAAT-ACTTTAGCGCAGTATGGAAGATTTTGCTC ************************ ****** **** *********** ********** *  U.I.T._2 T. _ c f . _co 1  TTGTGG-CAAGATCAAAGAGAGCCATTGGGG-TATTTCTATTTCATGGCGAGCGGTGAAA GCATGGGCAAGATCAAAGAGAGTCATTGGGGGTATTTCTATTTCATGGCGAGTGGTGAAA  *** *************** * * * * * * * * * * * * * * * * * * * * * * * * * * * *  *******  Figure 25 : T. cf. collaris and ft flavipes gut symbiont 2 ssu rRNA 92 % sequence identity is observed between the two sequences. Alignment was done using ClustalW and adjusted by eye. U.l.T. 2 is the R. flavipes gut symbiont 2. 82  U. I . T . _2 T ._cf ._col  TGCGTTGACCCATGGGAGAGAAACGAAAGCGAAGGCAAATACCCAAAGG-TTTCTGTCGA TATGTTGACCTATGGGAGAGAAACGAAAGCGAAGGCAAATACCCAAAGGGTTTCTATCGA * ******* ************************************** ***** ****  U.I.T._2 T._c f ._col  TCAAGGGCGAGAGTAGGGGGAGCGAACCGGATCAGAGACCCGGGTAGTCCCTACTGTAAA TCAAGGGCGAGAGTAGAGGGAGCGAACCGGATCAGAGACCCGGGTAGTCCCTACTGTAAA **************** *******************************************  U . I . T . _2 T . _c f . _co 1  CGATGTCGACAGGGGATTGTCTACTAGTTAGGCAGAACCTTAGCAAAAATGATAGTTCAT CGATGCCGACAGGGAATTGTCCACTTGGTGGACAGAACCTTAGCAAAAATGATAGTTC AT  *************************  ****************************  U. I . T . _2 T . _c f . _co 1  GGACTCTGGGGGAACTACGACTGCAAGGCTGAAACTTGAAGGAATTGACGGAAGGGCACA GGACTCTGGGGGAACTACGACTGC AAGGCTGAAACTTGAAGGAATTGACGGAAGGGC AGA  U. I . T . _2 T . _ C f . _COl  CCAGGGGTGGAGCCTGTGGCCTAATTTGAATCAACACGC^GGAAACTTACCAGAACCAGAT CCAGGGGTGGAGCCTGTGGCCTAATTTGAATCAACACGGGGAAACTTACCAGAACCAGAT ************************************************************  U . I . T . _2 T . _c f . _co 1  GATTCGATGACTTACATCTAAAAGAGCTTTAAGGATGGGTTTGATGGTGGTGCATGGCCG GATTCGATGACTTACATCTAAAAGAGCTTTAAGGATGGGTTTGATGGTGGTGCGTGGTCG * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * . * * * * * * * * * * **  U . I . T . _2 T . _c f . _co 1  TTGGTGGTGCGTGGGTTGACCTGTCACGAGTCGATTCAGTTAACGAGCGAGACTACCGCC TTGGTGGTCCGTGGGTTGACCTGTCACGAGTCGATTCAGTTAACGAGCGAGACTACCACC  *************************** * *.* **************************** *  ******** ************************************************  **  U . I . T . _2 T ._C f ._col  AAATATTAGCCAAGCT-ATACTTTAGTGGAAGCTTCTATTTGGGACTACCTGCGTCTA-G AAATATTAGCGATCCTTAGATTATAAGGAAAGCTTCTATTTGGGACTACCTGTTTTCAAG * * * * * * * * * * * ** * * * ** * * * * * * * * * * * * * * * * * * * * * * * * * * *  U. I . T . _2 T ._cf ._col  CGGGAGGAAGAGGGTAGCAATAACAGGTCCGTGATGTCCTTTAGATGCTCTGGGCTGCAC CGGGAGGAAGAGGGTAGCAATAACAGGTCCGTGGTGTCCTTTAGATGCTCTGGGCTGCAC  *********•••****•••*•***•********  **************************  U.I.T._2 T . _ c f._co1  GCGTGCTACAATGTGAATCA-AAAGGATTTAAAATCCGAAAGGATGCATTACTTCTTGAA GCGTGCTACAATGTGAATCACAAAGGAAAAATAAATCGAAAGAATGCATCACTTCTCC AA  U . I . T . _2 T ._cf ._col  AGATTCACGTAGCTGGGATTGACATTTGGAATCATTGTCATGAACCAGGAATCCCTTGTA AGATTCACGTAGCTGGGATTGACATTTGGAATTATTGTCATGAACCAGGAATCCCTTGTA ******************************** ***************************  U. I . T . _2 T. _ c f . _ c o l  AATGTATGTCAACACCGTACGTTGAATACGTCCCTGCCCTTTGTACACACCGCCCGTCGC AATGTATGTCAACATCGAACGTTGAATACGTCCCTGCCCTTTGTACACACCGCCCGTCGC * * * * * * * * * * * * * * ** * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *  U . I . T . _2 T . _c f . _co 1  TCCTACCGATTGGACGTTTTGGTGAAAATACTGGACTTCTTT -GAAAGGTATTTAA TCATACCGATTGGACGTTTTGGTGAAAATACTGGACTTTAATTTAAAGGAAGGTATTTAA ** * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * ***********  U.I.T._2 T._cf._col  ATCAGCTCGTCTAGAGGAAGGAGAAGTCGTAACAAGGTAACG ATCAGCTCGTCTAGAGGAAGGAGAAGTCGTAACAAGGTAACG ******************************************  ********************  ******  82a  * **  ****** ****** ******  **  Methanococcus voltae Giardia lamblia 95  Hexamita inflata ft flavipes gut symbiont 1 j— ft flavipes gut symbiont 2 Trichonympha agilis  100  100  99  - Trichonympha cf. collaris • Dientamoeba fragilis  98  - Metadevescovina extranea 96 100  Coronympha octonaria  93  Metadevescovina polyspira Tritrichomonas foetus  77  Trichomonas vaginalis 1 op__ jrichomitus trypanoides 83 80  40 37  Pseudotrichomonoides scroa Pseudotrichomonas keilini  >—100  Ditrichomonas honigbergi  0.1  Monocercomonas sp.  Figure 26 : Neighbour joining tree, Chapter Three The scale bar represents 10 substitutions per 100 positions. Distance matrix was made using a Kimura-2 parameter and a Tn/tv ratio of 2. The input order of taxa was jumbled.  83  Methanococcus voltae i — Giardia lamblia 90 1  — Hexamita inflata - R. flavipes gut symbiont 1 - ft flavipes gut symbiont 2 - Trichonympha agilis  100 89 100  -— Trichonympha cf. collaris - Metadevescovina extranea - Coronympha octonaria  96 96 73  - Metadevescovina polyspira - Dientamoeba fragilis  71  55  - Tritrichomonas foetus - Trichomonas vaginalis - Trichomitus trypanoides  100  34  60 46  - Pseudotrichomonoides scroa i — Pseudotrichomonas keilini  99  21  Ditrichomonas honigbergi Monocercomonas sp.  Figure 27 : Parsimony tree, Chapter Three Input order of taxa was jumbled 3x.  84  B  A  Figure 28 : Trichonympha cf. collaris probed with universal probe. A: Image viewed using phase contrast microscopy B: Image viewed using fluorescence microscopy  85  Figure 29 : Trichonympha cf. collaris incubated with the no-probe control. A: Image viewed using phase contrast microscopy B: Image viewed using fluorescence microscopy  86  S, strix  Figure 30: Trichonympha cf. collaris and $. strix incubated with specific probe. A: Image viewed using phase contrast microscopy B: Image viewed using fluorescence microscopy  87  Figure 31 : Trichomonas and S. strix probed with universal probe A: Image viewed using phase contrast microscopy B: Image viewed using fluorescence microscopy Cells labelled 1 are Trichomonas cells and the one labelled 2 IsStreblomastix.  88  A  B  Figure 32 : S. strix and Trichomonas incubated with the no-probe control A: Image viewed using phase contrast microscopy B: Image viewed using fluorescence microscopy Cells labelled 1 are Trichomonas cells and the one labelled 2 \sStreblomastix. Note the variability with which the Trichomonas cells autofluoresce. This caused difficulty in distinguishing positive hybridization results from background fluorescence. 89  Table 3: Sequencing primers used for Trichonympha cf. collaris SE0R2  249-  TTTAGGCTATTACGGG  SE0F2 SE0R3  1198-GGACACCACGGAACCT 940- GGGGTGGAGCCTGTGGC  SEQF3  424-  GATAGTCATTGTGTATC  Table 4 : Internal sequencing primers  These were used to obtain both strands of the full small subunit ribosomal RNA gene sequence. M13 forward and reverse primers were used as external primers to obtain the first round of sequence. At the beginning of each primer is the position to which it binds on the sequence.  90  Organism  Accession number  R. flavipes gut symbiont 2 Trichonympha agilis Trichonympha cf. collaris R. flavipes gut symbiont 1 Pseudotrichomonas keilini Pentatrichomonoides scroa Trichomonas vaginalis Ditrichomonas honigbergi Trichomitus trypanoides Metadevescovia polyspira Metadevescovia extranea Coronympha octonaria Monocercomonas sp. Dientamoeba fragilis Tritrichomonas foetus Hexamita inflata Giardia lamblia Methanococcus voltae  HSU17512 AB003920 AF023622 HSU17508 U17511 PS18SRNAG TJ17510 U17505 TT18SRR U17506 ME18SRRNA U17504 U17507 U37461 M81842 L07836 M54878 M59290  Table 5 : Organisms and accession numbers : Chapter Three The following 18 sequences were used to create the phylogenetic trees seen in Fig 26 and 27.  91  Chapter IV: Discussion Overview Both the specific and general questions introduced in Chapter One of this thesis can now be addressed. I will address the specific questions of where the hypermastigotes and the oxymonads place on the molecular phylogenetic tree by evaluating the results from Chapters Two and Three. This will show, within the reliability permitted by this study, what was the first sexual lineage to have diverged away from the eukaryotic line leading to the higher organisms. However, the general question to be addressed in this thesis is how and why sex first evolved. With the deepest diverging sexual lineage established, within the limits permitted by the available information, I can move on to infer some traits of the ancestral sexual cycle and look at the impact of those traits on the various theories of the origin of sex. This will lead to possible directions that future work might take in explaining our sexual past.  Oxymonad project Review of results The specific goal of the oxymonad project was to obtain a phylogenetic placement of the ssu rRNA gene sequence from Pyrsonympha and from this to infer the phylogenetic position of all oxymonads. Only 640 bp of sequence could be retrieved from a putative pyrsonymphid. This small amount of data is likely to contain a limited number of informative sites, which might account for the sequence not being reliably placed as a member of a specific clade. However, the putative pyrsonymphid sequence must have diverged after the parabasalans, due to the formation of a monophyletic clade by all sequences excluding the parabasalans and diplomonads (Fig 9,10). This result was obtained  92  using various phylogenetic methods and with 100 percent bootstrap values. From this it is also clear that the putative pyrsonymphid sequence is not closely related to the diplomonads.  Evaluation of Results Identification Due to the difficulty of culturing oxymonads, Pyrsonympha cells were manually isolated. However, cells could not withstand more than two rounds of washing and still remain intact. Mechanical methods of drawing liquid into the micropipet that cause less jarring suction might allow for more washing steps to be performed in future isolations. Also, removal of solution from the diluted gut fauna sample caused blurring in the field of vision under which the isolation was being observed. It was therefore possible that contaminant organisms were selected along with the targeted cells, and that the sequence obtained might not belong to Pyrsonympha.  Isolating the organisms under lower magnification would reduce  the blurring of the field of view, but this would also increase the risk of misidentification of the targeted cells. Several pieces of evidence argue against the putative pyrsonymphid sequence being derived from a contaminant. Clones with the identical restriction pattern to that containing the putative pyrsonymphid sequence were identified from independent PCR replicates. A search in a population of clones for contaminating sequences derived from identified gut protozoa was unsuccessful, despite the fact that the method for identifying contaminants was validated by the identification of both a bacterial and algal contaminant.  The identified gut protozoa from  Reticulitermes are mostly parabasalans. The putative pyrsonymphid sequence was excluded from the parabasalan clade and is unlikely to be derived from a trichomonad or hypermastigote contaminant. It is possible for a primer to select  93  against a specific sequence. However, the range of organisms amplified by these primers shows that they are not exclusive to a particular phylogenetic range. This makes it unlikely that Pyrsonympha was excluded on a phylogenetic basis. The evidence given here, and in more detail in Chapter Two, is not conclusive proof, as might be given by FISH studies. However, the most parsimonious explanation for the assembled data is that the putative pyrsonymphid sequence is actually derived from Pyrsonympha. Trees The bootstrap values for the exclusion of the parabasalans and diplomonads from the monophyletic clade formed by the higher organisms is very strong (Fig 9, 10). This shows that the parabasalan clade diverged before the putative pyrsonymphid sequence, wherever in the higher branches it may belong (Table 1). It must be pointed out that the ssu rRNA gene shows significant rate variation between sites in the gene and that this was not taken into account. However, an independent analysis of elongation factor alpha sequence data obtained by Moriya et al. (accession numbers AB007028 and AB007029) and analyzed by Andrew Roger (Fig 33) shows the same relative placement of the oxymonads to the parabasalans (Roger personal comm.). The placement of the parabasalan and diplomonad clades in the phylogenetic analysis in Chapter Two, independent of the relative placement of the oxymonads, is consistent with most other studies (Cavalier-Smith and Chao, 1996). Speculation on placement of the oxymonads A definite placement of the oxymonads was not possible at this time. Bootstrap values of 47% place the putative pyrsonymphid sequence with the percolozoan clade in the neighbour joining tree (Fig 10). In parsimony analysis of the non-bootstrapped data (not shown), the putative pyrsonymphid sequence also  94  diverged as a sister to the percolozoans. While this evidence is not nearly strong enough to be conclusive it is however, intriguing. Evidence suggesting similarity between the percolozoans and the oxymonads already exists. The discovery of Psalteriomonas lanterna (Broers et al, 1990) provides an example of a percolozoan with striking morphological similarities to the oxymonads. Its microtubular organizing region resembles the pre-axostylar ribbon of oxymonads. P. lanterna also has four basal bodies clustered together at the anterior end of the cell similar to the arrangement in oxymonads. P. lanterna and the oxymonads also share the lack of a Golgi and true mitochondria. These characteristics are likely adaptations to the anaerobic way of life. However, since this loss of the dictyosome and mitochondria due to an anaerobic lifestyle has occurred once in an established percolozoan, it is not unimaginable that the same process could have occurred in the oxymonads.  Evaluation of the morphological placement of the oxymonads The metamonads are classified together on their lack of characteristic organelles (mitochondria, hydrogenosomes, peroxisomes, and Golgi). Because losses can be secondarily derived and have been seen many times in other organisms this is not a strong basis for classification as was pointed out at the time of their grouping (Grasse, 1952), (Cavalier-Smith, 1987). P. lanterna also shares these characteristics, with the exception of speculations that there may exist modified mitochondria (Broers et al, 1990). It has also been speculated that the membrane bound cytoplasmic granules in Pyrsonympha may be hydrogenosomes or peroxisomes (Cavalier-Smith, 1987). Hydrogenosomes have also been postulated for P. lanterna (Broers et al, 1990). The best positive evidence for the classification of the metamonads seems to be the characteristic arrangement of the four basal bodies closely together. However,  95  P. lanterna also has its four basal bodies closely arranged at the anterior end of the cell (Broers et al, 1990). The MTOR in P. lanterna and the pre-axostylar ribbon in oxymonads (Brugerolle, 1991) have been speculated to be quite similar (Broers et al, 1990) and may be a shared trait. As summarized in Table 6, the morphological evidence for the placement of the oxymonads with the percolozoa (P. lanterna at least) is as strong, if not stronger, than with the metamonads. This placement is generally consistent with my distance analysis of the molecular data and should be pursued by obtaining more sequence data.  Speculation on evolution of the metamonads The phylogenetic analysis in this study indicates that the oxymonads and the diplomonads are not closely related. This means the oxymonads should be excluded from the metamonads. Information about the point of divergence of the retortomonads and their relation to the diplomonads is required before any further speculation on the evolution of the metamonads can be made.  Hypermastigote project Review of results The specific goal of the hypermastigote project was to phylogenetically place the ssu rRNA gene sequence of Trichonympha, inferred to be representative of all hypermastigotes. Trichonympha cf. collaris forms a monophyletic clade with Trichonympha agilis (Ohkuma et al, 1997) and R. flavipes gut symbiont 2 from the 1995 Gunderson et al study (Gunderson et al, 1995). These three sequences diverge as a sister group to the identified trichomonads, as might be expected for a separate class of  96  parabasalans. R. flavipes gut symbiont 1 diverges basally to the hypermastigote clade.  Evaluation of results Identification Numerous checks ensured that the Trichonympha cf. collaris sequence was derived from a Trichonympha species. The eukaryote-specific primers used in the PCR reduced the opportunity for amplification of bacterial contaminants. Isolation of Trichonympha cf. collaris from the entire gut contents was done by micro manipulation, which included both identification based on morphology and washing of the Trichonympha cells. PCR from an aliquot of the final washing step in the Trichonympha isolation yielded a very faint band of a different size than the one present in the Trichonympha PCR. This demonstrated that the isolation of Trichonympha was moderately but not completely successful, however cutting the Trichonympha band out of the gel distinguished between differently sized sequences. The BLAST results and phylogenetic trees show that the Trichonympha sequence is closely related to R.flavipesgut symbiont 2. The only protists common to the two gut environments from which the sequences are derived are Trichonympha species. The FISH data demonstrated that the specific probe binds to Trichonympha significantly above the background autofluorescence in the cells. The weak binding of the universal probe to Streblomastix cells means that the lack of binding of the specific probe to them can not discount Streblomastix as a source of the T. cf. collaris sequence. However, this result is also not inconsistent with the T. cf. collaris sequence being derived from Trichonympha. The fluorescence of wood in the Trichomonas cells also means that they could not be definitely excluded as a possible  97  source for the T. cf. collaris sequence on the basis of FISH data. However, Trichomonas sequence does exist, and the phylogenetic analysis of the T. cf. collaris sequence shows it to be quite divergent from that of Trichomonas. The final piece of evidence confirming the identity of the Trichonympha cf. collaris sequence was the existence of the Trichonympha agilis sequence submitted by Ohkuma et al. (Ohkuma et al, 1997). Their sequence has 92.3% identity to that of Trichonympha cf. collaris and forms a monophyletic clade with it and the R. flavipes gut symbiont 2 with 100% bootstrap values. Together the isolation method, phylogenetic analysis, and FISH data strongly suggest that the sequence I have obtained is from Trichonympha cf. collaris.  Trees Bootstrap values for the exclusion of the hypermastigotes from the trichomonad clade are 73% and 99% in parsimony (Fig 27) and neighbour joining (Fig 26) trees, respectively. This confirms the morphological assessment of the hypermastigotes as an independent class. The Trichonympha sequences form a monophyletic clade with R. flavipes gut symbiont 2 with high bootstrap support. The pairwise percent identity of R. flavipes gut symbiont 2 and Trichonympha agilis (93.4%) is higher than that of T. cf. collaris and T. agilis (92.3%). However, the latter two group together with high bootstrap values with both methods of analysis. This result was obtained regardless of whether gaps were counted, or whether highly mutated regions of the gene were eliminated. It is not, therefore, a product of the alignment but rather of how the algorithms treated the assembled data. Since the three sequences have high relative percent identities (1.1% difference between the two pairwise alignments), the algorithm may be sensitive to the relative percent identity of the sequences to other taxa in the analysis (Mable, personal comm.)  98  It has been recently shown that the microsporidia, a group thought to be ancient, based on both morphological (Cavalier-Smith, 1987) and ssu rRNA data (Vossbrinck et al., 1987), is actually related to the fungi. One interpretation of this result has been to question the validity of molecular phylogenetics based on ssu rRNA data. However, since the conclusions reached by ribosomal trees were corroborated by morphological comparisons, as well as data from other genes, I believe the problem with the microsporidia has raised serious questions about the reliability of phylogenetics in general. Rather than reject the entire field, I believe that the incident with the microsporidia underlines the importance of remembering that our studies are never conclusive and should always be re-evaluation when new information arises. The classification of the parabasalans as ancient is based on ssu rRNA trees (Cavalier-Smith, 1993), trees from HSP 60 (Roger, Clark and Doolittle, 1996) (which demonstrated the recent divergence of the microsporidians), as well as morphological studies (Cavalier-Smith, 1983). The classification of the parabasalans as ancient is fairly strong, with respect to other conclusions arrived at by phylogenetics.  Speculation on the evolution of parabasalans Until now the accepted scenario for the evolution of the parabasalans has been that the trichomonads were the most basal group and that the hypermastigotes were derived from within the trichomonad clade (Brugerolle and Taylor, 1977). Morphological similarities with the devescovinids suggested that the hypermastigotes were recently derived from that line (Fig 19). This entire scenario assumes that morphological simplicity was the ancestral state and that complexity increased through time in some of the lines. The trees that I have produced show the trichomonads as a monophyletic group, diversifying after the divergence of the hypermastigotes from the  99  trichomonad line (Fig 26, 27). This differs from the previous scheme in that the hypermastigotes are a deeply diverging line within the clade and not closely related to the devescovinids. The lack of flagella, axostyle, and basal bodies seen in Dientamoeba fragilis can now be confirmed as secondarily derived, as those traits are present in both the hypermastigote clade and trichomonad clade, within which the simple trichomonads diverge. Until the identity of R. flavipes gut symbiont 1 is established and other hypermastigotes are placed on the phylogenetic tree, it is not possible to determine whether the ancestral morphological state of the parabasalans tends towards simplicity or more complex versions of common traits. It is also necessary to obtain sequence from more hypermastigotes, since phylogenetic trees are sensitive to taxon inclusion. The topology of the tree may change once more hypermastigote sequences are included. This is apparent from the alternate divergence orders of the R. flavipes gut symbionts produced when the Trichonympha cf. collaris sequence was (Fig 26, 27) or was not (Fig 9, 10) included in trees. There is an obvious time discrepancy between the early divergence of the parabasalans and the late divergence of their hosts on the eukaryotic tree. Since several free-living trichomonads (Pseudotrichomonas, and Ditrichomonas) have been found in the anaerobic sediments of lakes (Farmer, 1993), this seems like a plausible environment in which the ancestral parabasalans may have lived prior to the evolution of their present-day hosts.  100  The ancestral sexual cycle and its inferred traits Deepest sexual line Due to their phylogenetic placement below the oxymonads, the parabasalans appear to be the modern descendants of the deepest diverging sexual lineage. It has been assumed throughout this thesis that sex has evolved only once and that all sexual organisms share a sexual cycle derived from the ancestral one. This means that the trichomonads for which there is no reported sexual cycle may be cryptically sexual or else secondarily asexual.  Hypermastigote sexual cycle There is variation in many aspects of the sexual cycle within the hypermastigotes. Observations on hypermastigote sexual cycles are taken from Cleveland's work cited here and reviewed by Grell (Cleveland, 1949), (Cleveland, 1965), (Grell, 1967). Only citations for specific exceptions will be referenced at the appropriate points. It is appropriate here to include a cautionary note on the work by Cleveland. In his extensive body of work, Cleveland describes sexual cycles giving detail of a variety of processes and physical states of the cells. His observations of minute details such as the stages of nuclear divisions and their correlation with cell processes, as well as observations of D N A replications could be regarded with a degree of skepticism since, for many of these processes, the distinctions are small and the techniques available to him might not have been able to allow him to distinguish them reliably. As well the genera assignments of organisms occurring within the same host (Cryptocercus in particular) may be suspect because of the difficulty in classifying such morphologically similar organisms. Underlining this problem is the fact that in the paper on the sexual cycle of Saccinobaculus (Cleveland, 1950), Notila is so similar to Saccinobaculus, that Cleveland refers to it  101  as the same genus. He then decides that it is different enough to deserve its own genus in his paper on its sexual cycle (Cleveland, 1950). This similarity could lead to mis-identification of genera. Processes might be attributed to cells of the wrong genera and therefore the order and even presence of events within a cycle might be incorrect. This is a problem in the oxymonads and to a lesser degree in the hypermastigotes, as the range of morphological diversity within the hypermastigotes is very large. Cleveland's descriptions of the gross aspects of the cells (number of flagella, presence or absence of axostyles and rostral apparati, or cytoplasmic granules) as well as gross cellular processes such as cell fusions, divisions, discarding of organelles and migration of nuclei should be regarded as reliable. These process are well within the scope of a meticulous worker such as Cleveland to have described using the techniques available to him. Similarly there is no reason to doubt his accounts of the timing of these processes relative to the life cycle of their hosts or the inducibility of the sexual processes. In my review of the sexual processes in the hypermastigotes I have dealt only with processes that I feel are reliably described by Cleveland. However, because the problem of mis-identification of genera could lead to a break-down of his observations, I feel that the verification of his work is very important. The correlation of the sexual cycle to the host molting cycle varies widely between hypermastigotes. In Trichonympha alone there is variation that seems correlated to the fate of the host gut lining. Species present in Cryptocercus punctulatus form cysts (Cleveland, 1949) and are ejected into the environment where they are then ingested by a new host. Other species do not form cysts but are instead transferred to new hosts in the increased transfer of gut fluid between termites that accompanies the time prior to their molting. Still others have become  102  asexual (Cleveland, 1965) in hosts where the gut lining is retained within the host hindgut after its molting. Ever since Cleveland described one-step meiosis (Cleveland, 1950), its significance and even validity (Haig, 1993) have been topics of debate. Although the majority of the hypermastigotes have two-step meiosis resembling that of higher organisms, both Leptospironympha (Cleveland, 1951) and Urinympha (Cleveland, 1951) are reported to undergo one-step meiosis. Because of this, and because the information regarding the process of one-step meiosis is so sparse, nothing will be inferred in this thesis about the ancestral state of one versus two-step meiosis. Full gamete fusion provides the cells with access to cytoplasmic factors and organelles as well as the nuclear complement. Although the majority of hypermastigotes undergo complete gamete fusion, Trichonympha magna (Cleveland, 1965) undergoes partial gamete fusion with the nucleus of one gamete migrating into the cytoplasm of the other. Because the order of divergence within the hypermastigotes is unknown, it is not possible to determine whether full or partial fusion is ancestral. Some characteristics are constant across the hypermastigotes, and shared by Mixotricha paradoxa. All have closed meiosis, where the nuclear envelope remains intact during nuclear division. Hypermastigotes also show varying degrees of gamete differentiation. In some species there is extensive physiological differentiation between gametes, in others barely any. In every case however, the gametocyte produces cells that behave differently during the course of the fusion and that must find a cell of the complementary type with which to fuse. Finally, hypermastigotes have rounds of asexual reproduction with an occasional round of sexual reproduction in response to the host molting hormone.  103  Ancestral sexual cycle Knowing the character state of three traits in the deepest diverging sexual line, I can now compare these states to those across the eukaryotic tree. This will allow me to infer the state of these traits in the ancestral sexual cycle.- The character state of traits in eukaryotic groups was reviewed in the Handbook of Protoctista (Margulis et al, 1993), J. D. Berger and F.J.R. Taylor's textbook for Biology 332 at UBC (Berger and Taylor, 1981), a review by Grell of Cleveland's work (Grell, 1967), and a volume of Parasitic Protozoa (Kreier and R., 1993). This version of the eukaryotic tree is based on a condensed version of the tree from the 1993 review by CavalierSmith (Cavalier-Smith, 1993), with newer information regarding the placement of the microsporidia (Germot, Philippe and Le Guyader, 1997), (Keeling and Doolittle, 1996), percolozoa (Cavalier-Smith and Chao, 1996), and morphological placement of the retortomonads as sisters to the diplomonads (Cavalier-Smith, 1981). The new information regarding the placement of the oxymonads and hypermastigotes described in this thesis is also taken into account. Of the three traits common to parabasalan sexual cycles, only the frequency of sex can be correlated with any reliability in the other eukaryotic taxa. As can be seen in Figure 34 almost all eukaryotes reproduce asexually, with occasional sexual reproduction which may or may not be in response to an environmental stimulus. I will refer to this type of reproduction as facultative, as opposed to sex every round of reproduction, which I will refer to as constitutive. Although constitutive sexual cycles do not exist for single-celled organisms they are theoretically possible. Such a hypothetical cycle could consist of mitotic divisions producing gametes, which then fuse, and undergo meiosis (Fig 35). From the pattern seen in Figure 34 then, it is most parsimonious to assume that the ancestral sexual cycle was facultative and that obligate or constitutive sexual cycles evolved only in the higher branches.  104  Implications for the origin of sex It is important to note that phylogenetic studies can only tell us about the state of the deepest diverging ancestor to leave modern descendants. If some time elapsed between the origin of sex and the last common sexual ancestor, any changes made in that time would not be detectable by this analysis. It has been noted that for many theories of the origin of sex, it is mathematically more favorable to have infrequent rather than obligate sex (Hurst and Peck, 1996). The implicit question with regards to these theories then was whether infrequent sex is biologically relevant and relevant to the ancestral population in which sex evolved. The fact that the ancestral cycle and many modern protozoan sexual cycles involve infrequent sex with intervening rounds of asexual reproduction may demonstrate a trade-off between the benefits and costs of sex. Costs, such as maintaining the ability to have sex, are incurred whether sex happens or not. However, other costs, such as the disruption of successful combinations of alleles, and finding a mate are only taken if sex occurs. It is well established that the greatest benefit of sex is reaped in a population of infrequent sexuals as compared with asexuals (Hurst and Peck, 1996). As sex becomes more frequent benefits remain constant but costs increase. The fact that sex appears to be an infrequent event in the ancestral cycle and in the current lower branches of the tree supports this mathematical result, and may indicate that a balance is reached where the benefits of occasional sex are enough to maintain the ability, but the costs each round prevent it from becoming constitutive. Some theories of the origin of sex require stringent conditions for the environment in which sex occurs and with respect to the type of sexual cycle. In a qualitative way, a facultative sexual system permits individuals to pay the full cost of sex only occasionally. This allows the benefits to continue outweighing the costs  105  but some of the restrictions set on the specific traits of the ancestors in obligate sexual models might be relaxed. The effect of the traits on various theories Parasitic elements causing the origin of sex The parasitic element theory is favoured when sex occurs every generation. This is not seen in nature and so the following restrictions must be set on the theory of the parasitic origin of sex. If a fitness cost of having the element is incurred regardless of whether sex occurs, then for the parasite to increase in the population, the total cost accumulated over the intervening asexual generations must be equal to, or less than, the spread of the parasite during the sexual generations. If the cost only occurs when sex does, then the frequency of sex is irrelevant. Defense against parasites as cause for the origin of sex From parasitic genetic elements causing the origin of sex, we move to the theory where a line of organisms evolved sexual reproduction in order to defend themselves against parasitic attack. Sexual reproduction is a complex physiological process with corresponding large fitness costs to some of its components. Additionally, some costs of sex exist regardless of whether sex occurs in that generation. If an external force, such as parasitic attack, is going to cause sex to evolve then the cost in fitness to organisms unable to escape the parasite due to their asexuality must also be large. There are three ways in which sex provides for increased defense against parasites, as laid out in Chapter One. For each one the benefit derived is greater, the more often sex happens since it is assumed that parasitic attack represents a constant danger. There must exist a balance between the benefits derived from using occasional sex to defend against parasites, and the costs incurred due, both to the parasites, and maintenance of the ability to have sex. A n analysis could be done to see how this balance can be reached. This would give  106  insight as to whether the theory proposing that sex allows for the defense against parasitic attack is relevant to the origin of sex. Freedom in a genetic environment: "Hitch-hiking" It has been shown theoretically that infrequent sex is effective at freeing a beneficial mutation from its genetic environment (Peck, 1994) and also at increasing the frequency with which beneficial mutations are brought together in a population (Green and Noakes, 1995). Qualitatively, sexual alleles benefit from genetic hitchhiking only for as long as the association with a favorable allele lasts (Barton, 1995). This means either that linkage between the alleles must be tight or beneficial mutations must be common. A gene that is loosely linked to a beneficial allele in a facultatively sexual organism is effectively much more tightly linked than in an obligately sexual one, and therefore will reap the benefits of associations with the beneficial allele for a longer time (Otto personal comm.). This favours facultative sex over obligate or constitutive sex, although asexuality is still the most favorable way to maintain linkage. Otto and Michalakis (Otto and Michalakis, 1998) point out that the rate of recombination (or in this case frequency of sex) could change in the population in response to the number of beneficial mutations. In a changing environment mutations that may have been neutral or deleterious may become beneficial. A mutant allele that increases the frequency of recombination in response to a change in the environment might be favoured. Mutation load reduction Of all of the types of theories, the set involving mutation load reduction has most frequently included analysis of facultative sexual cycles in its models. It is well established in these models (Kondrashov, 1985), (Pamilo, Nei and Li, 1987) that facultative sex provides equivalent evolutionary advantages to obligate sexuality.  107  As with all models of facultative sex, the question has been why sex is not therefore facultative. I have found that for the majority of eukaryotic lineages, and most likely in the ancestral cycle, it is. Capricious environments The facultative nature of the ancestral sexual cycle can answer many of the concerns raised about sex under fluctuating selection. This theory originally put forward by Williams (1966) and summarized by Bell (Bell, 1982) proposes that sex is advantageous in a changing environment as, under some conditions, it will produce more genotypes of higher fitness than would asexual reproduction. However, Maynard Smith (Maynard Smith, 1971) pointed out, that for sex to be beneficial enough to evolve, the environment must not simply be changing, but be changing in a capricious way. Similarly, current models show that in order for sex to be beneficial in a fluctuating environment, the epistasic interaction between alleles must frequently alternate between cooperativity and antagonism (Sasaki and Iwasa, 1987). In other words the epistatic interactions must also be capricious. Finally it has been shown that for obligate sex to evolve in a changing environment, the environment must be so harsh having therefore such a high selection coefficient on the population that the population is likely to be driven extinct (Kondrashov and Yampolsky, 1996). These harsh and capricious environments are not readily seen in nature and so the explanation that providing high fitness progeny in changing environments caused the origin of sex has been labeled as biologically insignificant. However, if sex occurred infrequently then it might allow the cells to be asexual during periods of environmental stability. Sasaki (Sasaki and Iwasa, 1987) examined the optimal recombination rate in fluctuating environments and determined that obligate sex was not favoured. Sex occurring only when induced by an environmental change would combine the benefits of sexuality with the  108  advantages of asexuality during times of stabilizing selection. Sewall Wright pointed out, as early as 1939, that facultative sexual cycles allows organisms to retain useful combinations of genes when it is appropriate, but break them up when needed (Wright, 1986). This would alleviate the need for frequent fluctuations and capriciousness in order to explain the evolution of sex in a changing environment. If sex occurs only in response to a change in the environment, then on the time scale of the sexual cycle, the environment is acting in a capricious way.  109  Final conclusions and unanswered questions This study has provided a look into our past. By finding the deepest diverging sexual lineage, I have inferred a possible trait of the ancestral sexual cycle. This has been used to examine some current theories on the origin of sex, and look to the future for what sorts of studies can be done to gain more insight into this problem. There are unanswered questions and loose ends to this thesis. The level of confidence of my results can be increased significantly with fluorescence in-situ hybridization data to identify the putative pyrsonymphid sequence. The current results can not demonstrate with certainty to which clade the putative pyrsonymphid sequence belongs, only that it does not diverge below the parabasalans. With more sequence data, the placement of the putative pyrsonymphid might be firmly established. Questions about the evolution of parabasalans, and various traits of the hypermastigote sexual cycle, can only be answered when more sequences from hypermastigotes are available. Knowing the deepest diverging sexual lineage allows for the ancestral state of traits in the sexual cycle to be inferred. This can be done once Cleveland's work is confirmed and was begun in this thesis by examining the frequency of sex. It has been shown for many theories of the origin of sex, that low rates of recombination favour its evolution. Just because something is mathematically favoured does not mean that it is evolutionarily relevant. My work shows that modeling facultative sexual cycles to explain the origin of sex is likely to be evolutionarily justified. In their 1996 paper, Hurst and Peck state, "The big theoretical problem seems to be... why don't more species reproduce asexually most of the time, with only occasional bouts of sex?" (Hurst and Peck, 1996). The answer is that they do. M y  110  findings show that modeling infrequent sex is the biologically and evolutionarily appropriate way to approach the question of how sex first evolved. This trend should be continued, and I have reviewed the effect of facultative sexual reproduction on some theories of the origin of sex. The search for answers to the question of how and why sex evolved is nowhere near completion. My work has shown a little about the path that our ancestors took in the evolution of sex and a little about the path that we should follow now to find out why.  Ill  Desulfuroc Sulfolobus -Thermoplas Halobacter Methanococ Tritrichomonas • Trichomonas Hexamita i diplomonad ^ G i a r d i a la Naegleria .  Blastocyst Oxymonadida Dinenympha  • Tetrahymen — Entamoeba •  Plasmodium  — — Euplotes c -Stylonychi Triticum a • Arabidopsi — — Acrasis ro Tetramitus  4  — Trypanosom Euglena gr Porphyra p —  Dictyostel  — Trichoderm — — Saccharomy — Absidiagl —  0.1  I  Onchocerca  Mus muscul IVI  MZ  Drosophila  Apis melli Figure 33:Maximum Likelihood tree of E.F.- a sequences This analysis was done by Andrew Roger and generously provided as a personal communication. With the exception of the parabasalans and oxymonads, shown in bold and given full genus names, the treefile and taxa names are unchanged from his file. 112  <  Diplomonads Retortomonads  Trichomonads  Fa  Hypermastigotes  Fa  Parabasalia  Percolozoa  N  Oxymonads Fa Leishmania  N  Trypanosoma N Euglena  Fa _ .  Entamoeba  N  Mycetozoa  Fa  r—  Ciliophora  Fa  inozoa  Fa  picomplexa  Fa  — Rhizopoda __ -J__  Fungi Fa RadiataCo/Fa  I- Bilateria Co/Fa  p-StreptophytaCo/Fa J I  L Chlorophyta Fa Haptophyta Cryptista  Lr  Fa  — Heterokonta  Fa  Florideophyceae  Fa  Bangiophyceae  Fa'  Figure 34: Simplified eukaryotic tree correlated with frequency of sex.  This tree is based on information from the literature and this thesis. Fa= facultative sex, Co= constitutive sex, N= no information. 113  Fa  114  Figure 35: Hypothetical obligate sexual cycle in a unicellular organism The haploid unicellular organism grows in cell size (A). The larger haploid cell now undergoes a DNA doubling (B), and mitotic gametogenic division (C) to produce two haploid gametes. These fuse (D) to produce a diploid zygote which undergoes two-step meiosis (E) producing four haploid cells. Each cell can now go through the cycle anew.  B  +  ?  +  +  No peroxisomes  +  +  Pre-axostyle  +  +  +  +  +  +»*  +  +  Four clustered noGolgi basal bodies  115  Table 6 : Morphological traits : metamonads and P. lanterna * No true mitochondria were found but modified mitochondria were postulated. **Membrane bound densely staining bodies were seen in Pyrsonympha. These may be hydrogenosomes peroxisomes or evidence of a Golgi.  +*  P. lanterna  +  Retortomonad +  +  +  +  Diplomonad  Oxymonad  No no mitochondria hydrogenosomes  name  Barton, N . H . , 1995. Linkage and the Limits to Natural Selection. Genetics 140:821841. Bell, G., 1982 The Masterpiece of Nature. University of California Press, Berkeley. Bell, G., 1993. The Sexual Nature of the Eukaryotic Genome. Journal of Heredity 84(5) :351-359. Berger, J. D. and Taylor, F. J. R., 1981 Principles of Protistology: The study of eukaryotic unicells as organisms. Unpublished, Bloodgood, R. A., Miller, K. R., Fitzharris, T. P. and Mcintosh, J. R., 1974. The Ultrastructure of Pyrsonympha and its Associated Microorganisms. Journal of Morphology 143:77-106. Broers, C. A. M . , Stumm, C. K., Vogels, G. D . and Brugerolle, G., 1990. PSalteriomonas lanterna gen. nov., sp. nov. a Free-living Amoeboflagellate Isolated from Freshwater Anaerobic Sediments. European Journal of Protistology 25:369-380. Brugerolle, G., 1991. Flagellar and cytoskeletal systems in amitochondriate flagellates: Archamoeba, Metamonada and Parabasala. Protoplasma 164:70-90. Brugerolle, G. and Taylor, F. J. R., 1977. Taxonomy, Cytology and Evolution of the Mastigophora. Proceedings of the Fifth International Congress of Protozoology. New York. 14-28. Buhse, H . E., Stamler, S. J. and Smith, H . E., 1975. Protracted maintenance of symbiotic polymastigote flagellates outside their termite host. Journal of Protozoology 22:11A-12A. Cavalier-Smith, T., 1981. Eukaryote kingdoms: seven or nine? Biosystems 14(3-4) :461-81. Cavalier-Smith, T., 1981. The origin and early evolution of the eukaryotic cell. Molecular and cellular aspects of microbial evolution. Cambridge. 33-84. Cavalier-Smith, T., 1983. A 6-kingdom classification and a unified phylogeny. Endocytobiology II. Walter de Gruyter, Berlin. 1027-1034. Cavalier-Smith, T., 1987. Eukaryotes with no mitochondria. Nature 326:332-333. Cavalier-Smith, T., 1993. Kingdom Protozoa and its 18 phyla. Microbiological Reviews 57(4) :953-994.  116  Cavalier-Smith, T., 1995. Cell cycles, diplokaryosis and the archezoan origin of sex. Archiv fur Protistenkund 145:189-207. Cavalier-Smith, T. and Chao, E. E., 1996. Molecular phylogeny of the free-living archezoan Trepomonas agilis and the nature of the first eukaryote. Journal of Molecular Evolution 43:551-562. Clark, C. G. and Roger, A. J., 1995. Direct evidence for secondary loss of mitochondria in Entamoeba histolytica. Proceedings of the National Academy of Science USA 92:6518-6521. Cleveland, L. R., 1949. Hormone-induced sexual cycles of flagellates I. Gametogenesis, fertilization, and meiosis in Trichonympha. Journal of Morphology 85(2) :197-296. Cleveland, L. R., 1950. Hormone-induced sexual cycle of flagellates.il Gametogenesis, fertilization, and meiosis in Oxymonas. Journal of Morphology 86:185-214. Cleveland, L. R., 1950. Hormone-induced sexual cycles of flagellates III, Gametogenesis, fertilization, and one-division meiosis in Saccinobaculus. Journal of Morphology 86:215-228. Cleveland, L. R., 1950. Hormone-induced sexual cycles of flagellates IV. Meiosis after syngamy and before nuclear fusion in Notila. Journal of Morphology 87:317-348. Cleveland, L. R., 1951. Hormone-induced sexual cycles of flagellates I. one-division meiosis and autogamy in Urinympha. Journal of Morphology 88(3) .•385-420. Cleveland, L. R., 1951. Hormone-induced sexual cycles of flagellates VI. Gametogenesis, fertilization, meiosis oocysts and gametocysts in Leptospironympha. Journal of Morphology 88:199-244. Cleveland, L. R., 1956. Brief Account of the Sexual Cycles of the Flagellates of Cryptocercus. Journal of Protozoology 3(4) :161-180. Cleveland, L. R., 1965. Fertilization in Trichonympha from termites. Archiv fur Protistenkund 108(8) :l-5. Cochrane, S. M . , H.E., S., Buhse, H . E. and Scammell, J. G., 1979. Structure of the attached stage of Pyrsonympha in the termite Reticulitermes flavipes Kollar. Protistologica 15:259-270. Collins, O. R. and Gong, T., 1985. Genetic relatedness of a former apomict and a heterothallic isolate in Didymium iridis (Myxomycetes). Mycologia 77(2) :300-307. 117  Corliss, J. O., 1994. A n interim utilitarian ("user -friendly") hierarchical classification and characterization of the protists. Acta Protozoologica 33:1-51. Cross, J. B., 1941. A study of Oxymonas minor Zeliff from the termite Kalotermes minor Hagen. University of California Publications in Zoology 43:379-403. Crow, J. F. and Kimura, M., 1965. Evolution in Sexual and Asexual Populations. The American Naturalist 99:439-450. DeLong, E. F., Wickham, G. S. and Pace, N . R., 1989. Phylogenetic stains: Ribosomal RNA-Based Probes for the identification of single cells. Science 243:1360-1363. Dyer, B. D., 1993. Phylum Zoomastigina Class Pyrsonymphida. Handbook of Protoctista. Jones and Bartlett, Boston. 266-269. Engels, B., 1992. Amplify 1.2B. Madison , Wisconsin, Farmer, M . , 1993. Ultrastructure of Ditrichomonas honigbergii N.G.,N. Sp. (Parabasalia) and Its Relationship to Amitochondrial Protists. Journal of Eukaryotic Microbiology 40(5) :619-626. Felsenstein, J., 1995. PHYLIP(Phylogeny Inference Package). Seattle, Department of Genetics, University of Washington. Germot, A., Philippe, H . and Le Guyader, H., 1997. Evidence for the loss of mitochondria in Microsporidia from a mitochondrial-type HSP70 in Nosema locustae. Molecular and Biochemical Parasitology 87(2) :159-168. Giovannoni, S. J., DeLong, E. F., Olsen, G. J. and Pace, N . R., 1988. Phylogenetic group-specific oligodeoxynucleotide probes for identification of single microbial cells. Journal of Bacteriology 170(2) :720-726. Grasse, P. P., 1952. Classe des Zooflagelles Zooflagellata ou Zoomastigina. Traite de Zoologie. Masson et Cie, Paris. 574-578. Grass6, P. P., 1952. Ordre des Pyrsonymphines. Traite de Zoologie. Masson+Cie, Paris. 789-800. Green, R. F. and Noakes, D. L. G., 1995. Is a Little Bit of Sex as Good as a Lot? Journal of Theoretical Biology 174:87-96. Grell, K. G., 1967. Sexual reproduction in protozoa. Research in Protozoology 2:147214.  118  Grosovsky, B. D. D. and Margulis, L., 1982. Termite Microbial Communities. Experimental Microbial Ecology. Blackwell Scientific, Oxford. 519-532. Gunderson, J., Hinkle, G., Leipe, D., et al., 1995. Phylogeny of trichomonads inferred from small-subunit rRNA sequences. Journal of Eukaryotic Microbiology 42(4) :4115. Haeckel, E., 1866  Generelle Morphologie der Organismen. Reimer, Berlin.  Haig, D., 1993. Alternatives to meiosis: the unusual genetics of red algae, microsporidia, and others. Journal of Theoretical Biology 163(1) :15-31. Hamilton, W. D., Axelrod, R. and Tanese, R., 1990. Sexual reproduction as an adaptation to resist parasites (A Review). Proceedings of the National Academy of Science USA 87:3566-3573. Hickey, D. A., 1982. Selfish DNA: a sexually-transmitted nuclear parasite. Genetics 101:519-531. Hollande, A . and Carruette-Valentin, J., 1971. Les attractophores, l'induction du fuseau et la division cellulaires chez les hypermastigines. Etude infrastructurale et revision sysb§matique des Trichonymphines et des Spirotrichonymphines. Protistologica VII:5-100. Hollande, A . and Caruette-Valentin, J., 1970. Appariement chromosomique et complexes synaptonematiques dans les noyaux en cours de depolyploidisation chez Pyrsonympha flagelata: le cycle evolutif des Pyrsonymphines symbiontes de Reticulitermes lucifugus. Contes Rendues de L'Academie Scientifique de Paris 270:2550-2555. Hurst, L. D. and Peck, J. R., 1996. Recent advances in understanding of the evolution and maintenance of sex. 11(2) :46-52. Keeling, P. J. and Doolittle, W. F., 1996. Alpha-tubulin from early diverging eukaryotic lineages and the evolution of the tubulin family. 13(10) :1297-1305. Kidder, G. W., 1929. Streblomastix strix, morphology and mitosis. University of California Publications in Zoology 33:109-122. Kirby, H . J., 1932. Flagellates of the genus Trichonympha in termites. University of California Publications in Zoology 37:350-454. Kirby, H . J., 1934. Protozoa in termites. Termites and Termite Control. University of California Press, Berkeley.  119  Kondrashov, A. S., 1985. Deleterious mutations as an evolutionary factor. II. Facultative apomixis and selfing. Genetics 111:635-653. Kondrashov, A. S., 1994. The asexual ploidy cycle and the origin of sex. Nature 370(6486) :213-216. Kondrashov, A. S. and Yampolsky, L. Y., 1996. Evolution of amphimixis and recombination under fluctuating seclection in one and many traits. Genetical Research, Cambridge 68:165-173. Kreier, J. P. and R., B. J., Ed., 1993. Parasitic protozoa. Academic Press Inc., San Diego, California. Lee, J. J., Hutner, S. H . and Bovee, E. C , Ed., 1985. Illustrated guide to the Protozoa. Society of Protozoologists, Kansas. Leipe, D. D., Gunderson, J. H., Nerad, T. A. and Sogin, M. L., 1993. Small subunit ribosomal RNA of Hexamita inflata and the quest for the first branch in the eukaryotic tree. Molecular and Biochemical Parasitology 59:41-48. Maguire, M . P., 1992. The evolution of meiosis. Journal of Theoretical Biology 154(1) :43-55. Margulis, L., 1996. Archaeal-eubacterial mergers in the origin of Eukarya: Phylogenetic classification of life. Proceedings of the National Academy of Science 93:1071-1076. Margulis, L., Corliss, J. O., Melkonian, M. and Chapman, D. J., Ed., 1993. Handbook of Protoctista. Jones and Bartlett, Boston, Mass. Maynard Smith, J., 1971. What Use is Sex? Journal of Theoretical Biology 30:319-335. Maynard Smith, J., 1978 The Evolution of Sex. Cambridge University Press, Cambridge. Muller, H . J., 1964. The relation of recombination to mutational advance. Mutation Research 1:2-9. Neefs, J.-M., Van de Per, Y., De Rijk, P., Chapelle, S. and De Wacher, R., 1993. Compilation of small ribosomal subunit RNA structures. Nucleic Acids Research 21(13) :3025-3049. Ohkuma, M., Ohtoko, K., Grunau, C. and Kudo, T., 1997. Phylogenetic identification of the symbiotic hypermastigote Trichonympha agilis in the hindgut of the termite Reticulitermes speratus based on small subunit rRNA sequence.  120  Otto, S. P. and Michalakis, 1998. The evolution of recombination in changing environments. Trends in Ecology and Evolution 13:145-151. Page, R. D. M . , 1996. TREEVIEW: A n application to display phylogenetic trees on personal computers. Computer Applications in the Biosciences 12:357-358. Pamilo, P., Nei, M . and Li, W.-H., 1987. Accumulation of mutations in sexual and asexual populations. Genetical Research, Cambridge 49:135-146. Pauling, L. and Zuckerkandl, E., 1963. Chemical Paleogenetics Molecular "Restoration Studies" of Extinct Forms of Life. Acta Chemica Scandinavia 17:S9-S16. Peck, J. R., 1994. A ruby in the rubbish: beneficial mutations, deleterious mutations and the evolution of sex. Genetics 137:597-606. Roger, A . J., Clark, C. G. and Doolittle, W. F., 1996. A possible mitochondrial gene in the early-branching amitochondriate protist Trichomonas vaginalis. Proceedings of the National Academy of Science USA 93(25) :14618-14622. Roger, A. J., Svard, S. G., Tovar, J., Clark, C. G., Smith, M. W., GiUin, F. D. and Sogin, M . L., 1998. A mitochondrial-like chaperonin 60 gene in Giardia lamblia: Evidence that diplomonads once harbored an endosymbiont related to the progenitor of mitochondria. Proceedings of the National Academy of Science USA 95:229-234. Ruppel, D. H . , 1978. Termites in British Columbia. Victoria, Environment Canada. Sasaki, A. and Iwasa, Y., 1987. Optimal Recombination Rate in Fluctuating Environments. Genetics 115:377-388. Smith, H . E., Stamler, S. J. and Buhse, H . E., 1975. A scanning electron microscope survey of the surface features of the polymastigote flagellates from Reticulitermes flavipes. Transactions of the American Microscopical Society 94(3) :401-410. Smith, S. W., Overbeek, R., Woese, C. R., Gilbert, W. and Gillevet, P. M . , 1994. The genetic data environment and expandable GUI for multiple sequence analysis. Computer Applications in the Biosciences 10:671-675. Strobeck, C , Maynard Smith, J. and Charlesworth, B., 1976. The Effects of Hitchhiking on a Gene for Recombination. Genetics 82:547-558. Thompson, J. D., Higgins, D. G. and Gibson, T. J., 1994. CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position specific gap penalties and weight matrix choice. Nucleic Acids Research 22(22) :4673-4680.  121  Tibayrenc, M . , Kjellberg, F., Arnaud, J., Oury, B., Breniere, S. F., Darde, M . - L . and Ayala, F. J., 1991. A r e eukaryotic microorganisms clonal or sexual? A population genetics vantage. Proceedings of the National Academy of Science U S A 88:5129-5133. Trivers, R. L . , 1972. Parental Investment and Sexual Selection. Sexual Selection and the Descent of Man. Chicago. 136-179. V a n Keulen, H . , Gutell, R., Gates, M . , Campbell, S., Erlandsen, S. L . , Jaroll, E . L . , K u l d a , J. and E . L . , M . , 1993. Unique phylogenetic position of Diplomonadida based on complete small subunit ribosomal R N A sequences of Giardia ardeae, G muris, G. duodenalis, and Hexamita species. The Faseb Journal 7:223-231. Vossbrinck, C . R., Maddox, J. V . , Friedman, S., Debrunner, V . B. and Woese, C . R., 1987. Ribosomal R N A sequence suggests microsporidia are extremely ancient eukaryotes. Nature 326(6111) :411-4. Wright, S., 1986. Statistical Genetics in Relation to Evolution. Selected Papers of Sewall Wright. University of Chicago Press, 328.  122  


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items