STUDENT ABSTRACTS

BIOLOGY

Labeling Studies on the de novo Biosynthetic Pathway of Lemna Minor

Alison K. Criss

The variation observed in the sphingolipid composition of plasma membranes is a function of the tissue from which the lipids are isolated and the organism in which they are synthesized. The activities primarily responsible for this diversity introduce hydroxyl groups or double bonds on the long-chain base sphinganine, as well as a hydroxyl on the acyl moiety of ceramides. However, the substrates preferred by the activities catalyzing these changes have not been conclusively identified. All sphingolipids contain a long-chain base moiety available for modification, and fatty acid hydroxylation may occur either before or after ceramide formation. To gain insight into these questions, the radioactive precursor [14 C] serine was provided to the duckweed Lemna minor and its incorporation into sphingolipids via the de novo biosynthetic pathway was observed over time. The labeling study revealed that desaturation of the long-chain base in plants is performed after its N-acylation to a ceramide, whereas its hydroxylation may preferentially occur once incorporated into a glucosylceramide (cerebroside). Evidence also suggested that nonhydroxyl fatty acids are utilized to make ceramide and subsequently hydroxylated. These results mirror the patterns of long-chain base and acyl distribution among the primary types of sphingolipids, i.e., free long-chain bases, ceramides, and glucosylceramides.

The labeling procedure was also utilized to examine the effects of the fumonisins, a potent family of ceramide synthesis inhibitors, upon the de novo sphingolipid biosynthetic pathway. Exposure of Lemna to 1 uM fumonisin B1, followed by incubation with [14C]-serine, resulted in the expected accumulation of the labeled, unmodified long-chain base sphinganine. However, the radioactivity incorporated into hydroxylated (phytosphingosine) and desaturated (sphingenine) long-chain bases was not altered. These findings provide further support for modification of long-chain bases in the form of ceramides or glucosylceramides and indicate that the increase in phytosphingosine levels previously observed with fumonisin is due to degradation of phytosphingosine-containing complex sphingolipids.

Characterization of Sphinganine Kinase in Corn Tissue

Gregory J. Crowther

Sphinganine kinase catalyzes the first step in the breakdown of the long-chain base sphinganine by converting it to sphinganine-1-phosphate. While previous studies of sphinganine kinase have been limited to animal systems, this thesis endeavored to characterize the enzyme in corn tissue. Enzyme activity was assayed by providing corn microsomes containing the kinase with radiolabeled substrate and quantifying the amount of radiolabeled product formed. Sphinganine kinase was found to be optimally active with D-erythro-sphingosine to some extent and that GTP and ADP were viable sources of phosphate for the enzyme. However, CTP and UTP apparently did not serve as substrate; neither did the long-chain bases DL-threo-sphinganine and D-erythro-phytosphingosine. Sphinganine kinase was largely unresponsive to changes in pH. A cell fractionation assay, in conjunction with previous work, suggested that sphinganine kinase is located in the endoplasmic reticulum, the Golgi apparatus, or the cytosol. Kinase activity was not significantly altered by the presence of the fungal toxin fumonisin.

A Survey of Amphibian Skin for Novel Antibacterial Peptides

Mark H. de Kanter

Antimicrobial peptides are highly conserved components of the innate defense systems in amphibians, insects, and mammals. Although these peptides can be grouped into three families based on their secondary structure, they share a common mode of cytolitic action through the disruption of microbial membranes. The abundance of these peptides in amphibian skin was first identified by Zasloff (1987) in Xenopus laevis, and subsequent research has isolated a variety of antimicrobial peptides from different amphibian species. Although our extracts from three tree frog species (Hyla cinerea, Litoria caerulea, and Osteophilus septentrionalis), the European common frog (Rana temporaria), and axolotl (Amystoma mexicano) failed to show antimicrobial activity on an Escherichia coli assay plate, a promising extraction technique using injections of noradrenalin was developed.

Characterization of in vivo dUTPase Activity in Cells Infected with Vaccinia Virus

T. Miko Enomoto

Vaccinia Poxviridae is a DNA virus which replicates autonomously in the cytoplasm of its host cell. Towards this end, vaccinia's genome encodes a number of distinct viral enzymes which perform functions analogous to hot cell enzymes. The activity of the viral enzyme often replaces that of its host cell analog during the course of infection, with the host enzyme being down-regulated as the infection progresses. Deoxyuridine Triphosphate Nucleotidohydrolase (dUTPase), which converts dUTP to dUMP and pyrophosphate, is thought to be one such enzyme. Assaying the in vivo activity of dUTPase in whole cell extracts is complicated by the presence of a competitive phosphatase which acts to convert dUTP to dUDP. This competitive phosphatase can be inhibited with low concentrations of EDTA. Profiles for both the host dUTPase and the viral dUTPase activity in vivo over the course of a vaccinia infection cycle in BSC40 cells are presented. Preliminary experiments to determine the phosphorylation state of both the viral enzyme and of the cellular homologue after infection are described.

Ecology of Benthic Protozoa and Rotifers on North Government Island

Jeremy Fox

I studied the mechanisms by which physical disturbance, trophic cascades, and interspecific competition interact to structure assemblages of benthic protozoa and rotifers in small rock pools on North Government Island in Lake Superior. My approach involved field observations, a field experiment, and a laboratory experiment.

Field observations examined the effect of storm wave wash on the pool communities. Pools low on the shore are washed most frequently by waves, and the frequency of wave wash decreases with increasing height on the short. The mass of benthic detritus in the pools increased with increasing height on the shore. Abundance of protozoa and rotifers per unit area covaried with density of benthic detritus per unit area. Protozoa were numerically dominant, with a characteristic ratio of protozoa and rotifers. This ratio varied with the height of the pool; rotifers were relatively less abundant in the uppermost pools. Both taxa were quite diverse at the genera level (15 rotifer genera, greater than 60 protozoa genera).

A field experiment examined the effects of tadpole and nutrient additions on protozoa and rotifers in mid-shore pools, using a crossed, 2 x 2 factorial design. Treatment of pools with added tadpoles, added nutrients, or both revealed an interaction between tadpoles and nutrients. At low (unmanipulated) nutrient levels, the presence of tadpoles enhanced protozoa and rotifer population densities. At high nutrient levels, tadpoles reduced these population densities.

A laboratory experiment examined the possibility that protozoa exclude rotifers by exploitation competition. The experiment utilized a crossed, 2 x 2 factorial design (+/- rotifer sp., +/- protozoan sp.). The rotifer Philodina sp. and the protozoan Paramecium multimicronucleatum were similar in terms of efficiency of bacterial grazing in batch culture, in the absence of competition. When cultured together, these two organisms exhibited an unstable equilibrium with no clear competitive dominant.

Results of the field observations indicate that the amount of detrital habitat available determines benthic protozoa and rotifer densities. Detrital habitat availability varies with frequency of wave wash and thus with location on the shore. Results of the field experiment indicate that, for pools of similar location, the interaction of tadpole (top down) and nutrient (bottom-up) effects determines variation in detrital habitat availability and protozoa and rotifer densities. The lab experiment demonstrates, however, that patterns of protozoa and rotifer distributions along the shore likely cannot be explained in a simple way solely by competition for food.

Habitat Preferences of Stream Invertebrates and their Responses to Experimental Dams

Andre Gerard

Given the basic (and well-demonstrated) assumption that many stream invertebrates are highly sensitive to the hydraulic characteristics of a given microhabitat, the building of small dams across a swiftly-moving section of stream should be expected -- through the sudden increase in water depth and decrease in velocity and turbulence -- to result in profound changes in invertebrate community structure. This experiment was designed to track how quickly and to what degree the invertebrate populations of a first-order stream would respond to such altered hydraulic conditions and whether a dammed habitat would be relatively attractive or relatively inhospitable.

Sections of the stream characterized as riffles (erosional habitats) or pools (depositional habitats) were used as controls against which responses of the experimentally dammed riffles (simulated depositional habitats) could be measured. Cinder blocks were used to convert riffles suddenly into simulated pools. The invertebrate populations of the three habitat types were sampled over time in a standard area, and the specimens from three sampling dates (one day prior to damming and then 4 and 24 days after damming) were identified to family (when possible) and counted to determine relative abundances. A total of 14900 animals (approximately 4415 per m2 of stream bed) in 37 taxa were analyzed.

With time, the dammed communities (which were originally identical to riffle communities) came to resemble the natural pool communities in several ways. Although riffles and pools were not significantly different in total population size, riffle communities typically had higher taxa richness, higher alpha diversity and evenness (Shannon index), and greater spatial variability (Morisita index) than pools; accordingly, the dammed riffles became significantly less taxa-rich, somewhat less diverse, and significantly less spatially variable after the dams were built. According to the Proportional Similarity Index, dams and pools had become as similar to each other by day 24 as dams and riffles had been prior to damming.

Twenty-four of the 37 taxa displayed a consistent habitat preference, as determined by significantly higher abundances in either riffles or pools; 21 preferred riffles and 3 preferred pools. Of the 21 taxa that preferred riffles, 10 - by a conservative estimate - responded negatively to the dams by day 24, and none responded positively. Five of the responding taxa had significantly declined in abundance even by day 4. Using more liberal statistical tests, an additional 7 taxa were found to have significantly declined in the dam habitats. None of the taxa that preferred natural pools either increased or decreased in abundance in the dammed sites during the course of this study.

Thus, the experimental dams altered the otherwise riffle conditions in a way that was sufficiently inhospitable that a significant component of the resident invertebrate communities responded quickly and unfavorably. That the dammed communities also grew progressively more similar to pool communities in both structure and taxa identity highlights the power of a simple environmental parameter to largely determine the organization of an entire community.

Phenotypic Plasticity in Chorus Frog Tadpoles as an Adaptive Response to Predation

Laura Good

This study examined the adaptive response of chorus frog tadpoles, Pseudacris triseriata, to their primary predator, the larvae of the dragonfly Aeshna juncea. Both species develop in small pools along the rocky shoreline of Isle Royale, Michigan. A competition-predation gradient occurs along the shoreline such that A. juncea live exclusively in pools high on the shore. P. triseriata lay their eggs in both high pools containing dragonflies as well as predator free pools lower on the shore. Therefore, the tadpoles regularly face a heterogeneous environment.

This study was comprised of two experiments which examined the tadpole's plastic response to this heterogeneous environment and its adaptive significance. Experiment I was designed to document the plastic response to the predator, both behaviorally and morphologically. This experiment also generated phenotypic variation which was used in Experiment II. The plasticity of the tadpoles continued to be measured throughout Experiment II in order to determine the window of plasciticity. The tadpoles were exposed to an actual predation event to ask whether selection was acting on the phenotypic variation. The first experiment in the study documented that the tadpoles respond differentially to the two types of environments. Tadpoles raised with restrained predators increased their time spent in inactivity, and developed higher tail fins. The results of Experiment II demonstrated that the tadpoles continued to respond morphologically to the presence of the predators, indicating that the window of developmental plasticity remains open throughout the larval period. Although no significant selection was found due to the small sample size, the data suggests that the tadpoles are selected upon in the same direction which their morphological response predicted.

This study illustrates that the plastic response of tadpoles to dragonfly larvae may be an example of adaptive plasticity. New findings include the late window for plasticity, the matching of individual tadpoles using photographic data, and the attempt to measure selection in a natural population.

Characterization of the Muscle Development Chain Unc-45 in C. elegans

Caroline Kim

Unc-45 is a muscle-affecting gene in Caenorhabditis elegans which appears to have a role in thick filament assembly. The first half of my project was a genetic confirmation of the null phenotype. The second half was to determine whether the exon locations predicted by the computer program Genefinder, using the genomic sequence provided by the C. elegans Genome Project, were correct.

To determine if st601, a lethal allele of unc-45, was a null mutation, the genotype of st601 over a deletion, wc5, was created. The examination of wc5/st601 heterozygotes corroborates the results of an earlier investigation which found a variety of lethal phenotypes. In particular, the lethal phenotype of embryos which were an amorphous clump of cells capable of muscle contraction without achieving the two-fold stage. It is too early to tell if we are observing the null phenotype. PCR of wc5 embryos to characterize the deletion and of st601 embryos to determine if it is a point mutation has not been successful, but it might resolve these issues in the future.

We have isolated and sequenced the 5' end of the gene using reverse transcriptase-PCR and cycle sequencing. We are confident that the 5' end of the unc-45 gene has been defined. An unc-45 clone, FN1S2-13.3 kb, found in prior microinjection studies to not rescue unc-45 function, was sequenced and shown to be missing the first Genefinder exon. We have also found a transpliced leader sequence, SL1, on the isolated cDNA. Sequencing efforts have progressed to covering up to the first half of exon 6. Overall, our sequence matched the Genefinder predictions.

Inhibition of Vaccinia Virus dUTPase Activity by 8N3ATP

Erica L. Mayer

Comparison of the sequence of vaccinia virus dUTPase with dUTPase sequences of other species has revealed the existence of five highly conserved motifs. It has been hypothesized that the active site of dUTPase is coded for by one or more of these motifs. Photoaffinity labeling, a technique in which an azido-linked nucleotide analog is used to label specific amino acid residues, will be utilized in the future to map the active site. Before this analysis can proceed, actual binding of the azido nucleotide analogs to the active site must be demonstrated. Protection binding experiments by N.A. Roseman and R.K. Evans suggest that both 8N3ATP and 5N3dUTP bind to the active site of vaccinia virus dUTPase. The purpose of my investigation was to demonstrate this binding in a biologically relevant manner through the use of kinetic analysis. The Km of the vaccinia dUTPase used in the assays was determined to be approximately 0.75 - 1.5 uM. Competitive inhibition of dUTPase activity by 8N3ATP was demonstrated, and the Ki of the 8N3ATP was calculated to be in the range of 94.6 -173 uM. Competitive inhibition requires a 15 minute incubation of enzyme and competitor prior to substrate addition, suggesting that the rate of dUTPase binding 8N3ATP is very slow. In inhibition experiments performed after crosslinking of dUTPase and 8N3ATP, the enzyme loses about one fourth activity post-UV exposure relative to an unexposed control. Yet the crosslinked enzyme demonstrates a further reduction in activity of about 10% compared to a non-crosslinked control. These dinetic results suggest that 8N3ATP binds the active site of dUTPase. Yet, achievement of inhibitory saturation of enzyme activity will be necessary for definitive proof before photoaffinity labeling of dUTPase can begin.

Yeast Artificial Chromosome Cloning in Tetrahymena thermophila

Heather McAuley

Tetrahymena thermophila is a ciliated protozoan which contains two distinct nuclei types, a germline micronucleus and a somatic macronucleus. the macronucleus is generated by the amplification and processing of the micronuclear DNA, and thus macronuclear clones can facilitate in the creation of a physical map of the micronucleus. The objective of this study was to develop a Yeast Artificial Chromosome (YAC) cloning procedure for the production of T. thermophila macronuclear clones. The following aspects of a multi-step cloning procedure were developed: the isolation, purification, and preparation of a pair of YAC cloning vectors, pJS97 and pJS98 (obtained from P. Hieter at Johns Hopkins University), the isolation of insert DNA from T. thermophila by the lysis of cells in low melting agarose (LMA), size selection of the insert DNA using pulsed field gel electrophoresis (PFGE), digestion of the insert DNA embedded in LMA with the restriction enzyme EagI, and the optimization of spheroplast transformations using the enzyme zymolyase. Progress has been made in all aspects of the procedure; however, much work still remains before macronuclear clones of T. thermophila can be obtained with efficiency.

Sexual Dimorphisms of Perception and the
Redevelopment of Autogenous Song Specificity in the Zebra Finch

Matthew T. Murrell

The song systems of many avian species are sexually dimorphic in that the females often possess similarly structured neural systems containing smaller nuclei (Nottebohm & Arnold, 1976). Perception of conspecific song in many passerine songbirds is one putative function of song system nuclei in females, who do not sing (Nottebohm, 1984; Brenowitz & Arnold, 1986; Brenowitz, 1986). Therefore, female zebra finches, following the developmental alterations in the song system nuclei that differentiate them from those of males, may utilize these song nuclei in the perception of conspecific song.

In the song system of many songbirds, including the zebra finch, neurons in the motor nuclei HVC and Ra respond to sounds the birds hear (Katz & Gurney, 1981; Margoliash 1983, Williams & Nottebohm 1985), and may be selective for stimuli as complex as the bird's own song (Doupe & Konishi, 1991; Margoliash, 1983, 1986, 1987; Margoliash & Konishi, 1985; Margoliash & Fortune, 1992). such song selectivity may also be important for song perception (Margoliash, 1983, 1986, 1987).

In the male zebra finch it was presumed that, following song development, the song crystallized and could not be altered. However, Williams and McKibben (1992) showed that damage to the tracheosyringeal nerve, which innervates the syrinx, induces plasticity in the song system and results in long-term alterations in the temporal patterning of song. Since these song changes follow approximately the same time course as normal song acquisition and learning, the residual plasticity may extend into the development of stimulus specificity for the bird's new song.

This experiment assessed the auditory response properties of neurons in HVC by presenting females with rearranged male songs and males with rearrangements of their own song. Females exhibited auditory response-sensitivity to the temporal sequence of song syllables but not to the orientation of the syllables comprising that sequence, whereas males' responses showed sensitivity to the orientation of the syllables but not the sequence of syllables comprising the song. Following right tracheosyringeal nerve damage, males displayed a retention of auditory response specificity in HVC for autogenous song, implying either that the plasciticity invoked following tracheosyringeal nerve damage extends into the redevelopment of neuronal response parameters, or that the postoperative song is directed towards a configuration that excites the preexisting HVC neurons with similar intensity as the preoperative song.

Site-Directed Mutagenesis of HIV-1 Reverse Transcriptase and Molecular Modeling of
dGTP Substrate Analogs

Max H. Nanao

Despite various attempts to curb the spread of AIDS, HIV-1, the virus responsible for AIDS, continues to spread at an alarming rate. Reverse Transcriptase (RT) is an enzyme necessary to the life cycle of the HIV, but exogenous to the human body. Therefore, RT represents an excellent target for a specific HIV inhibitor. to date, nucleotide based inhibitors such as AZT, ddI and ddC have been the most effective inhibitors of RT, illustrating the importance of understanding the detailed interactions of nucleotudes with this enzyme.

The first part of this three part study involved the characterization of interactions of dNTPs with a putative regulatory binding site by site-directed mutagenesis. Previous work suggested that binding of dNTPs to a specific domain on the p66 subunit (near Tyr- 318) was primarily electrostatic in nature and responsible for substrate inhibition (Utnam, 1993). Crosslinking studies using 8-azido-ATP, 8-azido-dATP and 5-azido-dUTP identified two dNTP binding sites on the p66 subunit; a tight binding domain near Tyr-318 and a weaker binding domain around Glu-28. Tyr-318 in particular appeared to be crosslinked by the photoactive probes. From the crystal structure of RT, the closest positive charge to Tyr-318 was a hisidine at position 96. Additionally, amino acid sequence alignments indicated that His-96 was highly conserved among reverse transcriptases. This conservation suggested that His-96 is important for catalysis or for regulation of RT. In this study, using site-directed mutagenesis, His-96 was changed to a neutral asparagine residue. It was found that neither the Kd of dNTP binding, nor the specific activity of the H96N mutant were significantly different from the wild-type enzyme, suggesting that His-96 is not involved in dNTP binding at the domain near Tyr-318.

The purpose of the second part of this study was to examine several nucleotide analogs, including 8-azido-dGTP, 8-amino-arabino-GTP, 8 bromo-dGTP, 8 amino-dGTP, dGTP, 8-azido-arabino-GTP and aragino-GTP through computer modeling. The potential energy as a function of the glycosidic torsion angle ([[chi]]) was calculated for each nucleotide analog in order to identify any energetic barriers of rotation. In A-form DNA, the glycosidic torsion angle of a nucleotide ([[chi]]), can vary from 10o to 40o, but on average is 20o (Ferrin et al, 1986). In B-form DNA, [[chi]] ranges from 40o to 90o and is on average 60o (Ferrin et al, 1986). This disparity in the final conformation of nucleotudes in a nucleic acid molecule could possibly be targeted in the design of a specific inhibitor of RT. Specifically, a nucleotide analog with an energetic barrier of rotation near [[chi]]= 60o could not be a substrate for the incorporation of nucleotides into B-form DNA, which is endogenous to the human body. By contrast, this nucleotide could be used as a substrate for a reaction catalyzing the incorporation of nucleotides into an A-form helix, such as a DNA-RNA hybrid, provided that the energetic barrier to rotation of [[chi]] was low near 20o. Having evaluated several dGTP analogs for these criteria, three nucleotide analogs appear to be good candidates for the selective incorporation by RT. Specifically, these modeling studies suggest that 8-amino-dGTP, 8-azido-dGTP and 8-azido-arabino-GTP could be a highly specific substrates for enzymes incorporating nucleotides into A-form helices, but comparatively poor substrates for the polymerization of B-form helices.

The third part of this study was the determination of equilibrium dissociation constants of several forms of RT for dideoxy-terminated primer/template. Wild type heterocimeric p66/p51, wild-type monomeric p66, M230A mutant heterodimer and the Y318T mutant heterodimer were studied. Both of the mutant proteins exhibited a significant decrease in binding affinity for primer/template, suggesting that their decreased catalytic capacities can be partially explained by weaker primer/template binding. Binding of the p66 monomer was 35.7 times weaker than the p66/p51 heterodimer, suggesting that the p51 subunit is important in the binding of primer/template.

Behavioral Effects of Mauthner Cell Double Ablation in the Goldfish (Carassius auratus)

Bridie C. Newman

Goldfish, Carassius auratus, respond to an abrupt acoustic stimulus with a rapid, C-type startle response (C-start) which is preceded by Mauthner cell (M-cell) activity in the goldfish. M-cells appear to initiate C-starts with other reticulospinal neurons contributing to the overall response. Electrolytic lesioning of the M-cell has demonstrated that goldfish are able to produce C-start responses in the absence of M-cells. Such non-Mauthner responses were no different in mechanical performance (angular velocity, displacement speed, etc.) or frequency of response than sham ablated fish. However, preliminary studies on M-cell ablation using a new procedure which minimizes damage to adjacent tissue have suggested that the M-cells may have a greater role in C-start response than currently attributed to them.

To further investigate the role of M-cells in C-start responses, both M-cells were selectively ablated in eight goldfish by locating their axon cap, moving a microelectrode 50um lateral to the cap, penetrating the M-cell soma and mechanically killing the cell. In six sham-operated control animals, the axon caps were located and the soma were not damaged. The probability of eliciting C-start responses and a series of kinematic parameters (i.e., latency, angle, velocity) were compared between control and experimental animals at three hours, one day, thirty days, and sixty days postoperatively. The success of the operations were confirmed morphologically at the end of behavioral testing.

Results confirm that C-starts with longer latencies can be initiated by non-Mauthner cells in the absence of both M-cells. Fish with double ablations demonstrated a significant decrease in the probability of response at three hours, one day, and thirty days, when compared to their respective controls. This reduced probability of response suggests that the M-cells may not only initiate C-starts, but also contribute to the excitation of other reticulospinal neurons involved in this type of startle response.

Characterization of the EXM2 Gene Via Transposon Mutagenesis

Tahira A. Palmer

A gene, termed EXM2 (exit from mitosis), which is involved in cell cycle control of budding yeast (Saccharomyces cerevisiae) was the focus of this study. Evidence suggests that the EXM2 gene may be checkpoint for exit from mitosis. The purpose of this project was to determine the functional limits and direction of transcription of the EXM2 gene via transposon mutagenesis. The bacterial transposon Tn10 LUK, which is a movable element of DNA containing the lacZ, kanamycin resistant, and UR3 gene markers, was induced to insert into a plasmid containing the EXM2 gene embedded in a large region of Saccharomyces cerevisiae genomic DNA. Experiments were performed that screened for plasmids that contained a Tn10 LUK insertion within the genomic region of EXM2. These plasmids were characterized via restriction enzyme analyses. Six plasmids that bear a Tn10 LUK insertion in the EXM2 genomic DNA have been identified and mapped. Presently, we are in the progress of screening for the phenotypic effects of these transposon insertions.

A Structure/Function Analysis of DeoxyuridineTriphosphatase from Vaccinia Virus

M. Adrian Rossi

Vaccinia Poxviridae is a DNA virus which replicates in the cytoplasm of its hosts. Since vaccinia does not depend completely on the host cell for replication, its genome encodes many of the required enzymes. One of these enzymes, dUTPase, catalyzes the hydrolysis of dUTP to dUMP. This reaction is important to the virus because the accumulation of dUTP in the host cell leads to the fragmentation of the viral DNA by the excision-repair systems of the host and virus. This reaction is also useful because the dUMP product enters the de novo synthesis pathway for dTTP.

Five conserved motifs have been identified by sequence analysis of the enzyme from eukaryotes and prokaryotes. The purification of a dUTPase mutant lacking motif 5 was attempted. The wild-type and the mutant gene were cloned into a pET14b plasmid which encodes for a 6-histidine tag that is transcribed to the amino terminus of the protein. This construct was then transformed into a bacterial over-expression system. Enzyme assays revealed that the histidine tag did not significantly diminish dUTPase activity. An affinity chromatography column directed at the histidine tag was used in an attempt to purify both the mutant and the wild-type proteins. The wild-type did not bind to the column, suggesting that its amino terminus was not accessible; the amino terminus may be buried within the three dimensional structure of the protein. The mutant was not recovered from the column, which is most likely due to damage to the column itself. Purification of the mutant from crude extract via gel filtration was also attempted, but the sample was too dilute to be detected by a Western Blot.

Isolation and Characterization of Suppressors of Two Temperature-Sensitive cdc 14 Alleles

Thad F. Schilling

The process of eukaryotic cell division has evolved into a series of highly regulated and tightly coordinated events, collectively termed the cell cycle (reviewed by Hartwell, et al (1974), Hartwell and Weinert (1989), Murray and Kirschner (1989)). Based on temporal changes in a dividing cell's morphology, four phases of cell division have been defined: G1, the gap before DNA replication; S, the DNA synthetic phase; G2, the gap subsequent to replication; and M, the mitotic phase. Progression through the cell cycle in most eukaryotes depends on the successful sequential completion of essential cellular processes including DNA replication, chromosome organization, centrosome duplication, spindle formation, chromosome migration, nuclear envelope degradation and reformation, and cytokinesis (Hartwell and Weinert (1989)). Many genes encode proteins which are essential in orchestrating or directly performing these complex processes in cellular division.

Effect on Mauthner Neuron Selective Axotomy on the
Startle Response of Goldfish (Carassius auratus)

D. Christopher Winters

Startle response frequency is significant attenuated, although not eliminated, in goldfish with bilateral selective axotomies of the Mauthner axons (p = 0.0001). The probability of eliciting a C-start remains significantly higher than would be expected in a fish with bilateral M-cell ablations (p = 0.0021). These data support the existence of a non-M-cell circuit that can perform C-starts in the absence of M-cell motor output. Additionally, these data suggest that the M-cells have an excitatory influence over the non-M network. Specifically, we propose that the M-cells exert this influence in one of two ways: either through direct or indirect excitatory connections, or through inhibition of inhibitory interneurons. Bilateral selective axotomy of the M-axons removes their excitatory influence and results in an attenuation of C-start frequency but not kinematic parameters.

CHEMISTRY

Synthetic Routes to the Azahexahelicenes

Nicholas P. Byrne

The azahexahelicenes are a series of large, nitrogen-substituted aromatic heterocycles. Five syntheses of three different azahexahelicenes were attempted. Most pathways relied on Wittig reactions to generate substituted ethenes, followed by photocyclization by UV light to generate fully aromatized products. The Siegrist reaction was also attempted. Steps were taken to avoid isomerization in the final photocyclization step. Problems were encountered with the radical bromination of the larger aromatic compounds and with the oxidation of 7-methylquinoline. Although no member of the azahexahelicene series was produced, the diversity of potential pathways to the azahexahelicenes makes it likely that future attempts will yield the desired product.

Further Work Toward the Total Synthesis of Paeoniflorigenone

Priscilla W. Carr

We have made important progress toward the elucidation of a viable synthetic pathway toward paeoniflorigenone. Paeoniflorigenone, a stereochemically and oxidatively complex monoterpene found in the roots of Paeonia albiflora, is known to exhibit anesthetic properties and is thought to be one of the pharmacologically active components of a drug used in East Asian folk medicine for the past 3000 years.

Former investigations have focused on the use of a common lactone intermediate to generate several members of a class of compounds found in the roots of P. albiflora. Recent research has focused on the development of a potential synthetic scheme toward paeoniflorigenone from said lactone intermediate. Former work has demonstrated the potential utility of a pathway that entails the following set of transformations of the lactone intermediate: benzoylation, reduction of the lactone moiety, deprotection to form a tetraol precursor and selective oxidation to form a tricarbonyl species that self-condenses to form paeoniflorigenone. While the earlier transformations had been performed successfully, the final oxidation and condensation had yet to be effected.

Herein, we report the results of a study with dual foci: the improvement of synthetic steps performed in recent years, and the putative observation of the desired final transformation, oxidation and condensation. Efforts this year have brought significant changes to the protocol for the reduction and deprotection reactions, with the results including decreased reaction times and increased yields. Finally, we have achieved limited success with the oxidation step of our synthesis, generating small quantities of a molecule that appears to be paeoniflorigenone.

Further Work on the Synthesis of Group (IV) Metallomesogens

Mark A. Cordes

Continuing with the work of previous researchers, the year's work has resulted in improved yields and more stream-lined workups of the intermediates on the route to the synthesis of group (IV) metallomesogens. The results of these pursuits have shown that yields of 75-90% are possible on each of the steps along the synthetic route to the ligand material. Although the synthesis of the ligands to be attached to the metal was quite successful, the actual attachment reaction involving a Grignard coupling reagent was not completed.

Initial probes into the group (IV) Grignard reactions using a simplified silyl analog of the silane ligand were very promising, giving the desired NMR shifts that had been shown in literature results to be consistent with a successful reaction. The later work with the full ligand appeared, from all physical evidence, (i.e., color, reaction rate, solid material formed, etc.) to have gone equally well, but this was not the case upon spectral analysis of the product mixture. The methylene group on the ligand which provided the NMR signal shifts necessary to determine the state of the reaction had disappeared in the case of the full ligand reaction. This surprising result shows that the Grignard complex attached to the methylene group had been quenched by a proton source and turned into a trimethyl silane instead of attaching to the group (IV) metal.

Quenching of Nitric Oxide (A2[[Sigma]]+, v = 0) Laser-Induced Fluorescence by Dimethyl Ether,
1-Propanol, Methanol, and Ethanol

Heather A. Cox

A 225 nm laser is used for single-photon excitation ( ) of nitric oxide. The fluorescence behavior for the transition ( ) is observed in the presence and absence of the quenching species dimethyl ether, ethanol, methanol, and 1-propanol at room temperature. Quenching rate cross-sections for these species are determined from analysis of the fluorescent lifetime of the mixture; they are 45 +/- 6.45 Å2, 94 +/- 6.7 Å2, 103 +/- 8.75 Å2, and 84 +/- 27.6 Å2, respectively. A hypothesis is ventured as to the cause of the irreproducibility of the 1-propanol quenching cross-section and qualitative predictions are made of the relative contributions of the harpoon, collision complex, and near-resonant electronic energy transfer quenching mechanisms to the quenching of nitric oxide by these species.

A cold cell was also designed, manufactured, and assembled for use in future experiments. Moreover, the experimental setup was adjusted to allow the introduction of liquid species with high vapor pressures into the system, which enabled the determination of the quenching cross-sections of these alcohols.

Generation of Novel Inorganic Radical Species and Their Detection
Using Electron Spin Resonance Spectroscopy

Rajiv Rajnikant Doshi

This project is concerned with the generation of Group IIIb to VIIb inorganic radical species and their detection through the use of electron spin resonance spectroscopy. Generation of these species involve the use of tetracyanoethylene, peroxydisulfate, and cerium (IV) oxidation systems. Methods such as steady-state continuous flow and spin trapping are used in the detection of these species. Novel species generated, such as the chlorite (ClO2), hypochlorite (ClO), and boron tetraphenyl (BPh4), are discussed.

The reactive intermediates involved in Grignard reagent formation of 1-bromohexylmagnesium bromide are analyzed with a spin trap, nitrosobenzene. Preliminary results indicate the formation of a reactive paramagnetic intermediate species, tentatively identified as an alkyl radical intermediate. Further studies are suggested to elucidate the exact nature of this species.

Synthesis of Novel Metallomesogens' Steps Towards the Preparation of Ni(CNR)4 Complexes

K. Jebrell Glover

We are interested in the preparation of novel metallomesogens. Particularly, this report is concerned with the synthesis of a new class of tetrahedral isonitrile metal complexes of the general formula Ni(CNR)4, where the R group is derived from a known mesogenic ligand. Through the study of these complexes we hope to probe the affect that tetrahedral geometry, low oxidation state (high electron density), R group structure, and intense color have on the already interesting optical and physical properties of liquid crystals.

Presently, we are working on a six step synthesis of the organic ligand, and the synthesis of Ni(COD)2, a Ni(CNR)4 precursor compound. To begin research on this new class of compounds an R group was chosen which is based on a p-alkoxybiphenyl (CN-C(CH3)2-(C6H4)2-O-(CH2)7CH3). When the compound is synthesized it will be checked for liquid crystalline properties using a heating stage, which provides a very controlled temperature environment, in conjunction with polarizing microscopy. Depending upon the physical character of this initial compound, the R group will likely be altered to enhance the liquid crystalline properties of the complex. These enhancements may include increasing the rigid or aromatic portion of the ligand, extending the region of conjugation, lengthening the terminal aliphatic chain, or adding permanent dipole functional groups.

Isolation and Characterization of Southeast Asian Dart Poison Components

Elizabeth A. Gray

The dart poisons used for hunting by indigenous peoples of southeast Asia are derived from the latex of the Antiaris toxicaria tree. The poisons are believed to contain cardiac glycosides as their toxic components. Members of this class of compounds are known to be potent inhibitors of Na+/K+ ATPase, a crucial enzyme which regulates the concentration gradients of sodium and potassium ions across cell membranes.

The primary focus of this work was to determine the structure/function relationships between the poison components and toxicity as measured by inhibition of Na+/K+ ATPase. To date, no systematic study has correlated structural features with bioactivity. By preparative high pressure liquid chromatography (HPLC) three individual poison components of intermediate polarity were isolated. All three components demonstrated the ability to inhibit Na+/K+ ATPase in an assay. Subsequent structural analysis of these components by proton nuclear magnetic resonance spectroscopy (NMR) revealed a signal in each spectrum which correlated with a proton on the unsaturated lactone of a cardiac glycoside. In addition, NMR studies identified an aldehyde proton signal in all three components, suggesting that the components belonged to the cardiac glycoside subclass having antiarigenin as the aglycone.

Synthesis and Isolation of 2,5-bis-4-tert-butylpyridinium-3,6-oxy-1,4-Benzoquinone and Attempted Synthesis and Isolation of Related Compounds.

Warren Grant Harbison, V

The halobenzoquinones chloranil, 1, and its ortho analog, 2, are excellent oxidizing agents in delicate organic reactions. Their cousin, 2,3-dichloro-5,6-dicyanobenzoquinone (DDQ), 3, is of limited solubility in many reaction solvents although it is very effective in dehydrogenations. An overall goal of this research is the synthesis of a class of pyridinium-substituted benzoquinone salts, such as 4, through use of chloranil and o-chloranil as substrates. The most recent extension of this exploratory process has been the synthesis of the dioxy-bis-4-t-butylpyridinium dizwitterion, 5. The pKa of 5 has been determined to be -3.3 according to the H0 acidity function. 5 is easily synthesized under atmospheric conditions and room temperature in dimethyl sulfoxide using 4-t-butylpyridine and chloranil as reagents.

An Analysis of the Catalytic Region of the Bacillus Subtilis
SOS Repressor

Jonathan B. Hargreaves

DinR is a DNA-binding protein involved in the regulation of an inducible DNA repair system in Bacillus subtilis. Under high pH conditions, DinR cleaves itself into two fragments of roughly equal size. The mechanism of this autodigestion is thought to involve a serine and a lysine residue, which are conserved in several regulator proteins in E. coli and bacteriophage lambda. The carboxy-terminal half of the molecule (one of the two cleavage products) contains both catalytic residues, and there has been some indication that this fragment retains proteolytic activity. The two cleavage products were separated by FPLC using a cation exchange column. The carboxy-terminal product demonstrated no activity when incubated with a general protease substrate, N[[alpha]]-benzoyl-DL-arginine p-nitroanilide, at high pH. In another attempt to isolate the DinR active site, a 16 kD fragment of DinR was generated by Clostripain proteolysis. The region of DinR contained in this fragment was identified by comparison of the amino acid sequence with SDS-PAGE and native gel electrophoresis results. The fragment contains both the substrate bond and the proposed catalytic residues. Under high pH conditions, it appears to undergo autodigestion. Because this fragment apparently retains catalytic activity, it is a useful subject for further study of the mechanism of DinR autodigestion.

Interactions of Two Putative B. Subtilis SOS Repressors with the
Regulatory Regions of Din Genes

Jennifer Katharine Hood

A 23-kD SOS repressor protein affinity-purified from Bacillus subtilis crude extracts (Lovett et al., 1993) and the product of the Bacillus subtilis dinR gene both bind specifically to the Bacillus subtilis consensus operator sequence, 5'-GAAC-N4-GTTC-3', and possess expected SOS repressor characteristics. The two proteins exhibit different binding behavior in mobility shift assays and Kd determinations, however, suggesting that they may be two distinct proteins. The DNA binding patterns of these putative repressors were investigated using mobility shift assays and hydroxyl radical footprinting. At saturating concentrations, DinR caused greater gel retardation of dinB and recA probes than did the affinity-purified repressor. Higher concentrations of DinR caused further retardation not seen with the affinity-purified repressor. Hydroxyl radical footprinting of the recA promoter region identified a putative region of DinR protection at nucleotides 38-45 and a similar but larger site of protection at nucleotides 38-52 with a crude extract representative of the affinity-purified repressor. These regions are upstream of the din consensus sequence and the flanking A-T-rich region. A possible site of protection within the consensus sequence at nucleotides 84 and 85 (both A-T base pairs) was also observed with the affinity-purified repressor, but not DinR. Further elucidation of these DNA-protein interactions may provide evidence of a Bacillus subtilis regulatory mechanism that is not present in the Escherichia coli SOS system.

Complex Dynamics in the Minimal, Contracted, and Extended Models
of the Alternative Pathway of Complement

Elizabeth Juang

Complement is a response mechanism of the immune system. Two pathways have been classified for complement, classical and alternative. The classical pathway is antibody mediated while the alternative pathway is not. Since the alternative pathway is independent of the presence of an infection, it is continuously active. A minimal model of the alternative pathway has been previously developed. The dynamics demonstrated by this minimal model correlates with expected physiological behavior and thus the model seems to be an accurate representation of the alternative pathway response.

In order to determine the model's validity, stability analysis must be carried out on the system. Stability theory utilizes the eigenvalues obtained from solving the characteristic polynomial of the Jacobian of the system to determine if a system is stable. A stable system is one that is independent of the initial conditions and parameters used. It is important that a biochemical system be stable because it exists in a web of interconnected responses, thus conditions are continuously varying. Since only solutions to polynomials of degree four or less are calculable, an effort to contract the existing six variable model to a four variable model must be attempted before applying stability theory. By applying a steady state assumption to the Michaelis Menten step in the minimal model, we contract the model to a five variable model. A purely mathematical comparison of the dynamics of the minimal and contracted models reveal that the two systems although different may not be incompatible. Both systems give rise to chaotic behavior at different parameter values. Further studies using physiologically relevant parameters in the contracted model are necessary to determine if it too is an accurate representation of the alternative pathway response. If so, simplification of the five variable model to a four variable model can be attempted and stability analysis applied.

The minimal model does not distinguish between host cells and infectious cells. An extended model of the alternative pathway of complement which distinguishes between these two types of cells is developed and its dynamics explored. Preliminary results using physiologically significant values indicate that the extended model accurately represents the expected dynamics in the self and non-self cells. In addition, a chaotic behavior has been found for a given set of parameter values. One drawback to the model is that the concentration of non-self cells is too high to be biologically relevant in these simulations.

Progress Towards a New Class of Organometallic Liquid Crystals:
Work Towards the Synthesis of Bis-cyclopentadienyl-bis-((4'-hexoxybiphenyl)-4'-dimethylsilylmethyl)titanium(IV

Jonathan R. Nitschke

Work towards the synthesis of bis-cyclopentadienyl-bis-((4'-hexoxybiphenyl)-4'-dimethylsilylmethyl) titanium(IV) is reported. In the past decade, many new metal-containing molecules with liquid crystalline phases have been synthesized and characterized. Many such compounds possess properties that make them very interesting from a materials scientist's point of view. Certain metal-containing liquid crystals, or metallomesogens, have been discovered to have liquid crystalline phases that span over 100 [[ring]]C. These highly polar metalloorganics also require less of an applied electric field to align them than the well-known purely organic liquid crystals. These two properties suggest metallomesogen-based display and sensor technologies might someday supersede their organic liquid crystal counterparts. Furthermore, entirely new technologies may be made possible by the electronic properties of the metal center. This work reports progress towards the synthesis of bis-cyclopentadienyl-bis-(4'-hexoxybiphenyl)-4'-dimethylsilylmethyl)titanium(IV) a molecule in which the ligands are tetrahedrally arranged around the metal center. We believe this substance may possess liquid crystalline properties. If it does, it will be the first known metallomesogen of tetrahedral coordination geometry. A test reaction has been successfully completed involving the lithiation of (CH3)3SiCH2Cl, and the reaction of the resultant (CH3)3SiCH2Li with Cp2TiCl2 to produce Cp2Ti(CH2Si(CH3)3)2, a simplified version of our target molecule. Synthesis of the 4-hexoxy-4'-(chloromethyldimethylsilyl)biphenyl ligand precursor has been completed; it remains to link this on to the titanocene center.

Evaporation Kinetics of Polar Stratospheric Clouds:
Ice and Nitric Acid Hydrates

Anne Nicole Normand

Polar stratospheric clouds (PSCs) serve as catalytic surfaces for reactions that activate compounds such as chlorine that eventually destroy ozone. I have analyzed data from a series of experiments that modeled PSC evaporation processes using thin films of ice and nitric acid hydrates. The latter included nitric acid monohydrate (NAM); nitric acid dihydrate (NAD); two forms of the trihydrate, known as [[alpha]]-NAT and ß-NAT; and amorphous mixtures of ice and HNO3 in 1:1 and 3:1 ratios. A final set of experiments involved ice with HCl. TPD and TPIR spectroscopy were used to monitor the desorption data, which were then analyzed using zero-order reaction kinetics and the Arrhenius equation:

rate = k = .

I calculated average values for the activation energy barrier, Ea, and for the preexponential factor A for each data set except the one with HCl, where averages are meaningless. For the various types of films, values range from 11.9+/-1.2 to 19.1+/-1.0 kcal/mol for Ea, and from 1031 to 1041 molecules/cm2sec for A. The calculated values agree with those reported in the literature; the fact that they all fall within experimental uncertainty confirms that our techniques provide an effective system to study desorption kinetics. Furthermore, the desorption profiles of the films seem to indicate the degree of order for a "crystalline" hydrated species within a particular temperature region. The kinetic parameters for these thin films should prove useful in generating better models of atmospheric PSCs and the reactions that lead ultimately to ozone depletion.

A Diels-Alder Route to Heteronaphthacenes

Daniel E. Patterson

Benzo[b]naphtho[2,3-d]furan 2a, benzo[b]naphtho[2,3-d]thiophene 2b, and N-methyl-5H-benzo[b]carbazole 2d were synthesized in one or two steps. Benzofuran, thianaphthene, and N-methylindole underwent Diels-Alder reactions with o-xylylene generated by the thermal decomposition of 1,3-dihydrobenzo[c]thiophene 2,2dioxide 14. The Diels-Alder reactions were performed in refluxing sulfolane, and in sealed ampoules without solvent. For benzofuran and thianaphthene, the adducts generated in the Diels-Alder reactions were the expected tetrahydro derivatives, which were aromatized to 2a and 2b, respectively, by treatment with sulfur in refluxing sulfolane. In the case of N-methylindole, the desired product was generated in the initial cycloaddition step.

Extensive research was also carried out on the generation of o-xylylene by the reaction of [[alpha]],[[alpha]]'-dibromo-o-xylene with tributyl(trimethylsilyl)-stannane. Additionally, isobenzofuran and [[alpha]],[[alpha]]'-dibromo-o-xylylene were investigated as dienes in reactions with the same dienophiles to synthesize the desired products.

Site-Directed Mutagenesis of Motif E in HIV-1 Reverse Transcriptase

Jean Marie Pesola

HIV-1 reverse transcriptase (RT) is a vital component of the HIV-1 virion and one of very few enzymes to be carried into a host cell by the virus. It is fundamental to the life cycle of the HIV virus because it catalyzes two crucial steps which lead to the integration of the provirus into the host genome: reverse transcription of its RNA genome into a single strand of DNA and the synthesis of a complementary DNA strand. The most effective AIDS treatments to date have been directed towards inactivating this enzyme, and such investigations occupy a prominent place at the forefront of current biomedical research. Nevertheless, many details of RT's function remain a mystery. The research presented here endeavors to contribute to a basic understanding of the activity of this enzyme.

We have used site-directed mutagenesis to probe the structure/function relationship in a small, conserved peptide loop of RT known as motif E. Two amino acid residues in this region were independently mutated: methionine-230 to alanine and tryptophan-229 to valine. These neighboring residues lie near the tip of motif E and proximal to the site of catalysis, the 3' hydroxyl group of the newly synthesized DNA strand. Both mutants have been partially characterized in an effort to discern their function. We report elevated Kms for dTTP relative to wild-type for both mutants, and a decrease in primer/template binding affinity for the methionine mutant. The most startling result is the finding that both mutations caused a loss of over 95% of polymerase activity. This tremendous loss of activity cannot be explained by the lower substrate binding affinities, which implies there must be a more important function for motif E.

A second project reported on here is the investigation of the effect of guanosine substrate analogs on the activity of wild-type RT. The two analogs examined, 8-azido-dGTP and 8-amino-dGTP, represent the first in a series of analogs which will be tested. Results indicate that while the azido compound has a lower binding affinity (Kd=17.4+/-3.6 uM) than the corresponding natural substrate, dGTP (Kd=2.4+/-0.6 uM), the amino compound has a higher binding affinity (Kd=1.20+/-0.06 u[[Mu]]). These preliminary experiments are setting the stage for a comprehensive exploration of compounds with characteristics that allow them to selectively inhibit RT, a requirement of any viable drug treatment.

Occurrence of Mercury and Chromium in Sediments and Fish
of Northwestern Massachusetts

Catherine E. Shawl

This thesis examines the global problem of heavy metal contamination using examples from northwestern Massachusetts. Chromium (Cr) is of particular interest because leather tanneries, potential point sources of this toxic metal, operated on the Hoosic River until the 1980's. Discoveries of high levels of toxic mercury (Hg) in water bodies with no known point sources also are increasingly common in this area. I examine concentrations of these trace metals in fish since consumption of contaminated fish is a primary pathway for uptake by humans. I also study concentrations in sediments because they are good indicators of transport and have sites for metal adsorption.

Microwave digestion techniques are developed to extract Cr and Hg from fish and sediment, and atomic absorption spectroscopy is used to quantify metal concentrations. My measurements of Hoosic River sediments and fish give no indication that the tanneries were point sources, that Cr has persisted in the environment, or that it is being taken up by fish in significant amounts. Hg concentrations in sediment are also within the background levels of this area, compared with sediment from Hopkins Memorial Forest. Hoosic River fish contain Hg in levels not thought to be hazardous to humans. Higher concentrations of Hg and Cr in larger fish provide evidence for the bioaccumulation of these metals. Metal concentrations vary with species of fish, suggesting that biomagnification and feeding habits are important considerations. A study of sediment from Eph's Pond shows that factors such as grain size, organic content, and presence of other metals are correlated with concentrations of Hg and Cr.

Targeted Catalysis: A Model for Preferential Enzymatic Digestion

Christopher S. Song

Experiments were performed to further elaborate and evaluate targeted catalysis as a model for preferential enzymatic digestion -- using the biotin-avidin system. Four distinct but related goals were addressed during the experiment: 1) To confirm the biotinylation of proteins via dot blot assays, 2) to detect actual complex formation between avidin and biotinylated molecules, 3) to further investigate the targeted catalysis model used by Wilchek and co-workers -- and tested by Tammie Smith `93, and 4) to begin work on a new model system involving cell lines. The first three goals were accomplished. However, the cell model system was not investigated. The dot blots were made to work, allowing for the confirmation of protein biotinylation. Complex formation was positively detected using electrophoresis and gel filtration chromatography. Finally, using biotinylated chymotrypsin and biotinylated transferrin, results were obtained that gave a slight indication of preferential cleavage of the target molecule over the free. A second enzyme system that was used did not further clarify targeted catalysis as a model system.

HIV-I Reverse Transcriptase: An Examination of Substrate Inhibition
with Site-directed Mutagenesis

Kirsten M. Williams

AIDS is an untreatable fatal epidemic, developed from infection with the Human Immunodeficiency Virus-I. HIV-I reverse transcriptase is both crucial and specific to the life-cycle of HIV-I. Thus, this enzyme provides a good target for AIDS treatments, since its inhibition would destroy the virus without detriment to the host cell. Current treatments are ineffective because of the low fidelity of HIV-I RT; the virus rapidly mutates, quickly creating resistant strain to all of the known inhibitors. Hence, the study of the structure/function relationship of HIV-I is critical, since a characterization of RT would significantly aid researchers in the development of a successful inhibitor. It is known that this heterodimeric enzyme, of a p66 and a p51 subunit, binds both primer/template and substrate to transcribe RNA into DNA. However, the details of the catalytic mechanism are largely unknown. Although the crystal structure has been published, it is not clear which amino acid residues participate in primer/template and dNTP binding; three aspartic acid residues, though, are believed to participate in binding two magnesium ions at the active site. This project attempts to determine the structure/function relationship of the finger region, specifically of alpha-helix A, using site-directed mutagenesis. Two mutations were introduced in the helix A of the p66 subunit: I37Q and K30M. K30M removes the positive charge from this position, while I37Q replaces a hydrophobic residue with a hydrophilic one; neither substitution alters the secondary structure of the helix. Enzyme activity was not greatly affected by the K30M mutation; the specific activity of the mutant was 72,300 units/mg while wild-type RT had 97,500 units of activity per mg of protein. However, both primer/template and nucleotide binding were significantly weakened in the K30M mutant. The Km for K30M was 31.0 uM, which was 6-fold higher than wild-type RT; and 3 times as much primer/template was required to reach saturation. These results suggest that alpha-helix A participates in binding the primer/template, partly through an electrostatic interaction to Lys-30. The results also indicate that nucleotide binding is greatly weakened, perhaps because the primer/template position is altered slightly, due to the absence of the positive charge at Lys-30. To address the influence of the K30M mutation on substrate inhibition of RT, assays were performed in the presence of inhibitory concentrations of dNTP substrates on K30M, wild-type RT, and another mutant of RT with an amino acid change near motif E(Y318T). Wild-type RT exhibited substrate inhibition of 2.3 fold at substrate concentrations of 400 uM and greater. The Kd approximated for the regulatory site was 175 uM for WT. In contrast to our expectations, K30M RT did not lose this regulatory function, in fact, given the increase in the Km, the identical inhibition value of 2.3 fold at 650 uM actually appears to represent increased inhibitory ability (since the effective substrate concentration is lower). However, no substrate inhibition was observed in Y318T. These results suggest that Tyr-318 is in a region of RT that is crucial for substrate inhibition. Further characterization of both the finger region and the region near Tyr-318 are necessary before the site of dNTP binding for substrate regulation can be identified. A better understanding of the structure/function relationships involving helix A and the site of substrate inhibition may be essential for advances toward the eradication of AIDS.

Strategies for Cloning Chimæras and Mutants of the
B. subtilis dinR
Gene Product

Donny Wong

The E. coli SOS DNA repair system is regulated by LexA and RecA. LexA functions as a repressor by binding the promoter regions of the damage-inducible genes of the SOS response, thereby blocking gene transcription by RNA polymerase. The RecA protein is activated in response to DNA damage, mediating the cleavage of LexA to de-repress the transcription of the damage-inducible genes of the SOS system. DinR is a homolog of LexA and functions similarly to repress an SOS-like response in B. subtilis. The RecA binding site on both LexA and DinR have not been characterized, though the carboxy terminal 36 amino acids are believed to be important for RecA-mediated cleavage.

Strategies for the cloning of chimæric proteins containing parts of both the lexA and dinR gene products have been outlined. The first chimæric protein, Din-Lex, would be a DinR protein in which the carboxy terminal domain is substituted with the corresponding region from the E. coli LexA protein. The second configuration (Lex-Din) is a LexA protein that contains the carboxy terminal domain from DinR. Successful products would be useful for structure-function analyses of the interactions between the B. subtilis and E. coli RecA proteins and the DinR and LexA repressors

Strategies for creating mutants of the B. subtilis dinR gene product by the Unique Site Elimination (USE) method of site-directed mutagenesis have also been outlined. Two mutants would contain a cysteine or an alanine residue in place of Ser127. Two other mutants would have Lys165 substituted with a methionine or an arginine residue. One other mutant would be a DinR protein truncated after 169 amino acids. Successful products would determine if DinR's conserved serine and lysine residues participate in RecA-mediated cleavage in a fashion similar to the proposed mechanism of cleavage of LexA.

COMPUTER SCIENCE

Finding Hamiltonian Cycles in Grid Graphs Without Holes

Stina S. Bridgeman

The Hamiltonian cycle problem is the problem of determining whether there exists a tour from one vertex in a graph back to the same vertex, visiting each other vertex exactly once. The problem is NP-complete in general, which raises the question of whether or not the problem is still NP-complete for various restricted classes of graphs. We describe progress on an algorithm to find Hamiltonian cycles in a highly restricted class of graphs known as grid graphs without holes.

TS++: Communication Specification Using Path Expressions in the
Tuple Space Programming Model

Stephen W. McLaughry

The tuple space model provides a simple, flexible means of specifying interprocess communication for parallel programming. However, this flexibility can result in incorrect programs, that are then extremely difficult to debug. We discuss a few extensions to the tuple space model, that will permit the programmer to specify information about the intended communication structure of a program. This will reduce the chance of error and simplify debugging tasks, without greatly changing the paradigm.

A Formal Analysis of C++: Type Rules and Semantics

Adam L. Seligman

C++ has rapidly become one of the most widely used programming languages in the world. The language is syntactically very similar to the C language, and in fact was originally developed as an extension to C. However, C++ is inherently object-oriented (OO) in paradigm, and therefore is a great departure from C. The various OO features supported by C++ are very powerful, but interact in complicated, and often surprising, ways.

In this thesis, we present a formal analysis of the core of C++, the method binding mechanism. This analysis will include features such as inheritance, overloading, overriding, hiding, and conversions. Our model will clearly explain method binding under all these features, without the ambiguity that colloquial descriptions can introduce.

Many of the intricacies of our formal model will demonstrate that the informal descriptions used to explain the language are not sophisticated or complete enough to explain certain examples. The type rules and operational semantics we present should provide the basis for a more accurate conceptual model of the language.

GEOLOGY

A Sedimentological and Stratigraphical Analysis of Glaciomarine Diamicts,
San Juan Islands, Washington

Michele Koppes

The San Juan Islands, in the northwest corner of Puget Sound, provide a unique locality where isostatic uplift has exposed thick beds of glaciomarine diamicts that were deposited in the late Pleistocene, as well as the sub-glacial tills that underlie them. Glaciomarine diamicts are good indicators of the calving margins of ice sheets, and of the retreat of these margins. The purpose of this study is twofold. The first is a detailed sedimentological and stratigraphic analysis of the diamicts in the San Juan Islands, to devise a depositional model indicating proximity to the ice margin and paleoenvironment of deposition. The second is the establishment of a set of criteria by which glaciomarine diamicts can be locally distinguished from related basal tills, in the absence of a fossil record.

Basing observations on models of inductive lithofacies created by Domack (1982), Smith and Hunter (1989) and Eyles et al. (1985), stratigraphic analysis showed that the substrate and the conformity of the diamicts' lower contacts determined proximity to the margin. By mapping diamicts and pebbly silts at eleven exposures around the islands, I determined that interbedded and/or conformable diamicts represented proximity to or contact with the ice margin, pebbly silts indicated proximity to the margin but distance from high energy meltwater tunnels, and massive and/or graded diamicts implied distance from the margin. The relative distances were based on the inductive processes of rain out and slumping down the slope of the submarine fan, settling out of the suspended fine load, and iceberg rafting, respectively. Upward coarsening of clast size and the loss of stratification within the diamict facies also indicated increasing distance from the ice margin, as the primary depositional processes shifted from rain out and meltwater channel deposition to iceberg rafting.

A comparison of glaciomarine diamicts and sub-glacial tills provided textural and clast axial ratio characteristics by which the two types of sediments could be distinguished. I collected and performed sieve and hydrometer analysis on 18 samples of diamicts, pebbly silts and tills, and measured clast orientation and axial lengths at 8 locations. Results showed that tills contained larger fractions of fine sand, very fine silt and clay, whereas glaciomarine diamicts contained higher percentages of coarse and medium sand. Diamicts also showed greater variation in clast shape.

Late Pleistocene Rocky Shorelines and the Record of Sea-Level Change at Punta Chivato,
Baja California Sur

Laura Libbey

A wide range of features representing rocky shorelines active during oxygen isotope substage 5e (120-130,000 years ago) are preserved on Punta Chivato and Playa La Palmita, Baja California Sur, Mexico. This locality fulfills the prediction of Hayes et al. (1993) that rocky shores in the Gulf of California and their attendant information can be preserved in the recent rock record. On the Punta Chivato promontory, a wave-cut terrace is preserved in andesite cliffs at an average elevation of +7.9m. Evidence for a virtually monospecific reef 71m long at an elevation of +7.75m, and a surge channel are also retained on the promontory. On nearby Isla Santa Ines, the remains of another reef and terrace system are recognizable.

Adjacent to Punta Chivato at Playa La Palmita, six sub-environments are preserved: (1) a 4m-high paleo-sea cliff; (2) a 61m-long intertidal abrasion platform composed of hydrothermal dolomite; (3) an oyster-capped hard ground; (4) a surrogate rocky shore composed of storm-transported cobbles and boulders; (5) a paleo-island; and (6) a tidal shoal. The quality of preservation of fossils associated with all of these sub-environments is exceptional. At the paleo-sea cliff, which is presently +4.25m to +6.25m above sea level, 15 species of encrusting organisms are preserved intact and in life position. This represents the highest number of encrusting species preserved on any known substage-5e rocky shoreline, based on a global survey of 50 rocky shores dating from the last interglacial epoch. The 12,500m2 paleo-island is elevated +11m above sea level and the 2600m2 tidal shoal rests +7.75m above sea level. Fossils retained in growth position along the margins of these features show distinct ecological zonation based upon prevailing wind direction and wave energy, and most bivalves are preserved in the life position. Again, this study represents the only known reference to a substage-5e paleo-island or tidal shoal. Lastly, although many articles regarding substage-5e rocky shorelines discuss intertidal abrasion platforms, this study is unique in its description of a platform composed of hydrothermal dolomite associated with a paleo-hot spring. The platform is extensively colonized by boring bivalves.

During substage 5e a global transgressive-regressive cycle occurred. The record of this cycle is preserved at Playa La Palmita above the rocky shorelines as a succession of silty limestones, a storm bed, clean limestones, and a regressive bed. The quality of preservation and the diversity of the fossils contained in these layers, which belong to the Mulegé Formation, is outstanding, with 77 species present. The organisms represent a sandy intertidal to subtidal environment with scattered rocks embedded in the soft substrate. Coralline algae is by far the most abundant organism in these beds; the organic limestones of this locality are composed predominantly (45%-85%) of coralline red algae. The transgressive-regressive cycle deposited 2, 018, 740 m3 of carbonate sediments on Playa La Palmita.

Playa La Palmita is located on the leeward, southern side of the protective Punta Chivato promontory, which buffers it from the strong prevailing northerly winter winds. The resulting sheltered environment is the dominant influence on the quality of preservation and diversity of organisms at Playa La Palmita. The findings of this study suggest that researchers seeking ancient rocky shores in the rock record should search along the protected sides of promontories, headlands, or peninsulas for deposits of the same caliber as those at Playa La Palmita.

The Classification of Marine Transgressive Holocene Sand Bodies Found in Seismic Profiles of Long Island Sound

Sarah Mills

Through the use of high-resolution seismic profiling in Long Island Sound, the stratigraphy of post-Pleistocene deposition can be interpreted to contain five major facies: glacial sediments, lake clays, channel-fill deposits, deltaic deposits, and recent marine sediments. The contact between the channel-fill deposits and the modern marine and deltaic sediments marks the time of eustatic sea level rise, which can be seen in the transgressive marine unconformity as a high-resolution reflector within the seismic profiles.

Directly overlying the unconformity and underlying the recent marine sediments are sand bodies that range in height from 0.70 to 10.23 m, in width from 50 m to 2.5 km, and in shore-parallel length from 100 m to 5 km. These sand bodies exhibit an asymmetry that includes a long shoreward width and a short seaward width, separated by a break in slope. The underlying unconformity slope ranges between 0.0000 and 0.0018.

The sand bodies identified in the seismic profiles of Long Island Sound are categorized into two morphologic clusters that relate to modern features. The first cluster contains sand bodies with an average height of 2.6 m. These are either: (1) beach deposits, or (2) offshore bars, based on their sizes and morphologies. The features in the second cluster have an average height of 6.5 m, which is indicative of the following modern feature dimensions: (1) barrier islands, or (2) spits.

These sand bodies are also clustered into three groups based on average depths of formation. These depths are: 19, 28, and 37 m. Sand bodies formed at these three depths during times of sea-level standstill or during times of slow sea-level rise. Sea level then rose at rapid rates, causing the burial of the sand bodies by recent marine sediments. The preservation of the sand bodies during sea level rise was dependent on their large sizes, the rapid rate of sea level rise, the protective nature of Long Island Sound, and a high sediment supply. The three depths of sand body formation are indicative of three standstill events followed by three rapid episodes in sea-level rise.

Map analyses of sand body deposition conclude that patterns of ancient deposition mirror the stages of sea levels during standstill events. Patterns of deposition are evident along: (1) the north shore of Long Island where there was extensive reworking of glacial outwash deposits, and (2) the south shore of Connecticut, near river mouths, where glacial and terrestrial runoff sediments added to the formation of sand bodies.

A Morphological and Geochemical Comparison of a Partially Reclaimed Open Mine Pit to a Similar but Undisturbed Site, Cooke City, Montana

Michael Montag

This study compares an abandoned mine site near Cooke City, MT, known as the Como Pit, which was strip mined for gold and copper during the early and middle parts of this century, and then partially reclaimed within the past two years, to a nearby area on Scotch Bonnet Mountain that is quite similar but was never mined. The regional geology is dominated by three felsic sulfide-rich stocks, which caused large amounts of mineralization and concentration of valuable metals such as gold, copper, silver, and lead, and now is the cause of acid rock drainage leading to ferricrete formation. The area, known as the New World Mining District, has been actively mined for these metals since the late 1870's, and is the site of the New World Gold Mine proposed by Crown Butte, Inc., which is still in the process of environmental review at the time of this writing.

I made geomorphic observations to describe the many erosion features found in the Como Pit, which ranged from wash zones to super gullies. I then computed an estimation of sediment loss in the Pit over the last year from transects across gullies that I measured at every 15m of elevation change, calculating that at least 1850m3 of sediment was lost from the approximately 15 hectares of the Pit. This is equivalent to a loss of a layer of sediment 1.2 cm thick for every hectare in the Pit over the last 11 months.

Analysis of local water geochemistry shows that the abandoned mine is definitely impacting the waters of the area. By providing a comparison between the two sites, as well as comparisons of water geochemistry along a given watercourse, and using a combination of field measurements and laboratory work, I was able to determine that the pH of water flowing from the Como Pit is significantly lower than that flowing at Scotch Bonnet, and has significantly higher concentrations of the heavy metals: iron, copper, and zinc. Analysis for anions such as SO42- showed that there were significant sulfides in solution in this sulfide rich area.

Soil geochemistry, measured from samples taken from 21 pits spread over the two sites, showed that soil pH did not appear to be controlled by the nearby disturbance of the area. Loss on ignition analyses simply showed that the upper portions of the soil profile tend to have more organic matter in them, and the percentage of organic matter was not affected to any great degree by mining. An acid extraction to determine leachable heavy metals content shows that the disturbance probably significantly increases the total levels of heavy metals in the soil near the mine site. The data also suggest that the mine has at least some effect on elevating concentrations of heavy metals in the upper portions of the Scotch Bonnet profiles as well, since concentrations of heavy metals tend to be high in the upper portions of the soil profile at both the Como Pit and at Scotch Bonnet. This could be the result of adsorption due to varied size of sediment, organic fixing, or even dust blowing from the barren mine site.

My geomorphic and geochemical data demonstrate that mining has had a significant impact on this high alpine to subalpine ecosystem. There are quite obvious differences between the Como Pit and Scotch Bonnet in erosion rates and gully formation, as well as water geochemistry. Also, soil geochemistry is sufficiently abnormal to believe that some of the results observed are as a result of human surface and hydrologic disturbance.

The Metavolcanic Wall Rocks of the Gouldsboro Pluton: Dyer Neck and Vicinity, Maine

John Phipps

The lower Paleozoic metavolcanic rocks of the Coastal Volcanic Belt, extending from north-eastern Massachusetts through Maine and into New Brunswick, consist of primarily bimodal assemblages of mafic and felsic volcanic rocks. Large, well-preserved sections have been studied in Penobscot Bay and the Cranberry Islands to the southwest and in the Machias-Eastport area to the northeast, but very little work has been done in the area in between. This study discusses isolated sections of mafic and felsic flows, pyroclastics and dikes, locally intruded by fine grained diorite, occurring adjacent to and within the Gouldsboro granite and a rapakivi granite on Dyer Neck and as thin septa between the two granites on Gouldsboro Neck.

A section of mafic flows covering about a square kilometer lies on Rogers Point on the west side of Dyer Neck, continuing westward across The Narrows to Garden Point on Gouldsboro Neck. Toward its southeastern end it is intruded by fine grained diorite which itself is intruded by the Gouldsboro granite. The volcanics appear to be overlain by a polymict conglomerate that is poorly exposed at the southern edge of the volcanics section, and which may correlate with the locality 10 kilometers to the west described by Gilman and Lash (1988). The metabasalt is cut by both mafic and silicic dikes and retains little primary structure as a result of amphibolite grade metamorphism and considerable shearing. The one remnant of original structure consists of discontinuous zones of amygdules, which suggest that the original layering may dip to the south. About six kilometers south of Rogers Point, at Lobster Cove, there is a roof pendant of metavolcanics in rapakivi granite. This steeply dipping exposure is about 60 meters thick and is comprised of 30 meters of basal pyroclastics containing flattened lithic fragments and fiamme, an equigranular felsic tuff, a 10 meter layer of metabasalt, and a capping felsic tuff. This exposure preserves more primary structure, but units within it are thoroughly recrystallized and metamorphosed to amphibolite grade.

Major and trace element geochemistry performed by XRF and INAA techniques provide apparently conflicting information about the tectonic setting of emplacement for these metavolcanics. They plot on tectonic discrimination diagrams as having either island arc or mid ocean ridge affinities. These contradictions between subduction associations and rifting associations are reconciled by recent studies that suggest subduction zones often exhibit some extensional characteristics. Based on these models, correlations with other rocks in the Coastal Volcanic Belt, and geochronological data, a model is proposed in which a composite Avalonian terrane approaches North America across the remains of the Iapetus Ocean with subduction zones dipping oppositely under each of these crustal blocks. Conveyance was accompanied by extension in parts of Avalonia throughout the Silurian and collision of the two terranes occurred in early Devonian in the Acadian orogeny.

Stress Perturbations Near Active Faults: Applications to Southern California

Demian Saffer

The goal of this study is to formulate a viable elastic model that explains the observed stresses and secondary fault patterns near major transcurrent faults. Classical faulting theory predicts that s1 will be oriented at approximately 30o to faults. Recently collected field data near the San Andreas fault, California (Mount & Suppe, 1987; Zoback et al., 1987) indicate that s1 in the horizontal plane is rotated to an angle of 85o with the fault trace locally, and remains oriented at approximately 50o to the fault trace in the far field. This pattern of s1 orientation has since been observed near the Great Sumatran fault, the Dead Sea Transform, the Philippine fault, and the Alpine fault, New Zealand (Mount & Suppe, 1992). In order to reconcile high regional deviatoric stress with measured stress orientations, previous workers have concluded that the San Andreas fault must be a zone of low shear strength. While this is a valuable qualitative insight, a quantitative model that explains the rotation of s1 near crustal weak zones has been lacking.

Previous attempts to model transcurrent faults have treated fault boundaries between two crustal plates as cracks embedded in a single elastic body. While this approach is valuable for the case of intraplate faults, it is conceptually inaccurate for the case of plate boundary faults. Because transcurrent faults break both the upper and lower surfaces of the elastic upper crust, they are best viewed as interfaces between multiple elastic bodies in contact.

I formulate an elastic stress model that describes the stress field surrounding a single fault boundary containing a discrete weak segment in horizontal planes of a given depth. The complete stress field is expressed as the sum of a uniaxial tectonic compression, the standard state stresses, and a stress perturbation caused by slip on weak fault segments. The stress perturbation decays exponentially with distance from the fault plane, and variations in stress along strike are described using a Fourier cosine series. The specific cases of extensional and compressional fault bends are examined by superimposing the stress fields surrounding two offset weak segments.

This model predicts s1 orientations in horizontal planes that match closely with field measurements. In the past, researchers have argued that because s1 remains normal to the San Andreas fault trace through its compressional bend near the Garlock fault, the San Andreas fault must be weak along its entire length. The results of this model demonstrate that identical s1 orientations can be caused by two offset weak segments with a locked zone between them, suggesting that the San Andreas fault need not be weak along its entirety to account for observed s1 trajectories. In addition, the predicted map areas susceptible to secondary faulting and the predicted orientations of ancillary faults mimic observed fault geometry. The model also predicts the formation and continued faulting on buried or "blind" thrust faults.

Foundering of the Pliocene Santa Inez Archipelago in the Gulf of California:
Baja California Sur, Mexico

Maximino Simian

A Pliocene archipelago of 5 small islands is preserved in the Punta Chivato region and adjacent Bahia Santa Ines. Marine deposits attributed to the San Marcos Formation are dated as early to middle Pliocene in age on the basis of echinoderm index fossils: Clypeaster bowensi Weaver (1908), Dendraster grandi and Encope sverdrupi Durham (1950). Island cores are composed of Miocene volcanics belonging to the Comondú Group and skirted by transgressive/regressive carbonate ramps sloping at angles of 6o from elevations of approximately 100 m to present sea level. The radial trends and consistent dips of these ramps indicate no tectonic tilting took place in Pliocene and later times. The islands were partly affected by foundering, because the eustatic rise in Pliocene sea level was only 25-35 m. A set of associated lithologies was deposited on the island shorelines, including interfingered basal conglomerates, siltstones and carbonate pocket beaches. A diverse biofacies with gastropods and bivalves occurs in the basal conglomerate and limestone of all islands, while a distinct coral facies is found only on the north shore of the Punta Chivato promontory. Sand dollars and oysters are found only on its south shore. The distribution of these energy-sensitive facies can be related to past circulation in the Gulf of California. Predominant winds from the north stimulated the development of corals in an exposed setting, while sand dollars and oysters thrived in a more protected setting on the west and south shores. These Pliocene deposits indicate the Gulf of California, though somewhat smaller, had developed with well-established circulation patterns and a biological diversity comparable to today.

Cascade Volcanics on the Southern Flanks of Brown Mountain, Southern Oregon

Peter Taylor

Brown Mountain is a 7,311 ft. andesitic cone standing to the southwest of Mt. McLoughlin, a typical composite volcano of the Cascade Range. Below the peaks of these two cones lie a diverse range of volcanics characteristic of the Cascade volcanic province. The Cascade Range is a product of arc volcanism related to the subduction of the Juan de Fuca Plate beneath the North American Plate. However, in southern Oregon, the Cascades abut the western reaches of the extensional Basin and Range province, a region known for primitive basaltic volcanism. This seven square mile study is part of a larger research project investigating what relationship, if any, these two volcanic regimes may have in production of extrusives throughout the southern Cascades.

With the exception of the recent Brown Mountain Andesite (~50,000 years old), the units of this field area hover about the 3 Ma mark. This age date is noted as the starting point for increasing levels of basaltic volcanism within the larger area south of McLoughlin (Mertzman, 1993). The northern sections of the field area are covered by the blocky andesite flows of Brown Mountain Andesite. To the south, outcrops of the older units form the summits of several hills amidst the drainage patterns of Little Butte Creek. Cox Butte summit, source of Cox Butte Basaltic Andesite, is the lone source vent in this area.

All units are holocrystalline and porphyritic. Phenocrysts consist mainly of plagioclase, olivine, orthopyroxene, and clinopyroxene. Glomeroporphyritic clumping is pronounced in several units. Cinder Pit Basalt is an olivine-phyric unit with large (3-5mm) euhedral olivine phenocrysts aligned in a flow oriented ground mass.

Geochemical data classifies the six units into distinct rock types: one basalt, two basaltic andesites, and three andesites. All units are calc-alkaline and may thus be related to the subduction of the Juan de Fuca Plate to the west. Study of geochemical trends against age reveals what may be two series of volcanics, each related by similar differentiation trends. An initial andesitic series shows maturation into Esther Applegate Two-Pyroxene Andesite by approximately 3.25 Ma. The second basaltic series initiates with Cinder Pit Basalt at approximately 2.9 Ma - after the 3 Ma mark cited for increased basaltic activity. Continued increase of basaltic activity is evidenced by the presence of high-alkali olivine tholeiites (HAOT), an extrusive related to extension, as early as 2 Ma to the south and east of my area. Additionally, the epicenters of the recent extensional Klamath Falls earthquakes of 1993 were located within the eastern Cascade margin. Both pieces of evidence indicate possible movement of the Basin and Range province westward into the Cascade Range.

This westward motion may be caused by a decrease in subduction rates since 7 Ma. In a subduction regime in which the subducting slab sinks towards the mantle after subduction, a slowing of rates would increase the overall dip of the slab. Thus, the zones of partial melting, arc magmatism, and back-arc rifting would move back towards the subduction trench.

MATHEMATICS

Finding Incompressible Surfaces

Jeffrey Bevelander

In an attempt to supplement the machinery already developed to identify knots and links, this thesis presents an algorithm that will detect the presence of incompressible surfaces, which can assist in the process of knot and link identification and analysis. Examining the link in its ideal triangulated form, the algorithm produces the simpler surfaces that can be expressed "nicely" by interconnecting triangles and quadrilaterals. Expressing a surface "nicely" basically means that each triangle and each quadrilateral used in the representation lies entirely in a single tetrahedron, and that none of the polygons intersect each other in a given tetrahedron. These two restrictions create a combinatorial requirement for a surface that sits "nicely" in the tetrahedra. Utilization of this requirement allows for the identification of the sets of triangles and quadrilaterals that lie "nicely" within the tetrahedra, and upon associating these sets of polygons with the surfaces they form, it is then possible to identify the surfaces that can be placed "nicely" into a link's ideal triangulated form. Since, incompressible surfaces always have a "nice" representation in the ideal triangulation, once we have identified our "nice" surfaces, we will have found all the simple incompressible surfaces in a knot complement, and gathered information that will further our efforts towards identifying the knot or link.

Analysis of Manifolds Using Morse-Smale Homology

David DelaCruz

A topological object can be defined as a set of points in a space. A particularly beautiful topological object to examine is the manifold.

Given two n-manifolds, it is interesting to know if they are topologically equivalent -- that is, whether we can bend, twist, and stretch one to get the other. The topological invariants of two manifolds must be the same if two manifolds are the same under topology, one such invariant is homology groups.

Singular homology theory describes homology groups for manifolds in the abstract. Morse-Smale homology allows us to capture homology groups algebraically as well.

The thesis is an exposition of singular homology theory, Morse theory, and Morse-Smale homology, followed by an algorithm that can be used to approximate Morse Smale gradient flows on algebraically defined manifolds.

Computing the Intersection Homology Groups of a Complex Algebraic Variety

Christopher French

An algorithm is given to compute the intersection homology groups for a complex algebraic variety. Two previously developed algorithms, the Collin's cad algorithm and Prill's Adjacency algorithm, are presented and used.

One Approach to Factoring Multivariate Rational Polynomials Over the Complex Numbers

Dimitry Korsunsky

Factoring a given multivariate polynomial is an important task in symbolic computation. Potential uses for an efficient solution to this problem could be found in various branches of applied mathematics, such as computer-aided design and theorem proving. Several algorithms giving different methods for factoring multivariate polynomials had been created over the years (Noether 1922, Davenport and Trager 1981, Christov and Grigoryev 1983, etc.). The theoretical basis for the algorithm, implementation of which is the subject of this paper had been put forward in Bajaj et al [4]. There it is proved that the suggested approach when implemented in parallel will execute in shorter time as compared to earlier solutions. Although a sequential solution, which was implemented, achieves lower efficiency there is a significant advantage in that it can be used in a large variety of settings. The program had been written using the Mathematica software package.

Bumper Drawings: A New Type of Proximity Drawing

Michael Pelsmajer

Open and closed ß-bumper drawings are defined exactly as ß-drawings, except that proximity regions contain no vertices and edges. Open and closed [[infinity]]-bumper graphs are completely classified. Maximal outerplanar drawings are defined for 1-bumper drawings, and partial results and methods are detailed.

Hyperbolic 3-Orbifolds

Edward Welsh

Suppose we have a hyperbolic 3-orbifold with discrete fundamental group G generated by two elliptic transformations, alpha and beta. This thesis finds restrictions on the possible distances between the axes of alpha and beta under various conditions. Introductions to both hyperbolic geometry and algebraic topology are included.

PHYSICS

Laser Cooling and Trapping of Rubidium Atoms

Patrick R. Frierson

Significant improvements have been made on the design and operation of the laser and magnetic field coils necessary for the stable operation of a magneto-optical trap for Rb. An old piezo that resonated at 400 Hz and led to instability in the frequency of the main laser for cooling and trapping was replaced with one that resonates at 900 Hz. The electronics have been optimized for the lock boxes that hold the frequencies of the main and re-pumper lasers stable. Currently, stabilities of better than 1 MHz are possible for the main laser. Finally, the Helmoltz coil configuration has been improved so that it can completely cancel stray field in the region of the MOT.

Measuring the Doppler Effect in the Hydrogen Maser

Christopher B. Kim

We present an experiment to test an analysis of the Doppler effect in the hydrogen maser. Catherine Hirshfeld and Craig Epifanio have developed a density matrix perturbation theory of motional narrowing in the hydrogen maser, and we continued the experiment conducted by Epifanio. We modified his pulsed magnetic resonance experiment to enable us to account for signal drift and a residual oscillating moment induced by the pulsing, and we have enhanced the data analysis through the successful implementation of computer programs that model radiation damping in the hydrogen maser. Analysis of results shows promising, though not conclusive, agreement with the theory.

Construction of a Background Free Signal Averaged Autocorrelator

Kira E. Maginnis

This thesis reports the construction and testing of a signal averaged, background free, autocorrelator. The instrument was constructed to measure the duration of laser pulses with a wavelength between 1.3 and 1.55 mm, and a maximum pulse duration of 200 picoseconds. The autocorrelator is designed in two separate parts: the physical apparatus necessary for generating a second harmonic signal, and a computer control interface. Testing on the computer interface has been completed with favorable results, as have tests on the physical apparatus. Testing the completed autocorrelator awaits the availability of a pulsed laser of known duration.

Quantum Cloning

Colin F. McCormick

It is a well-established result that information contained in quantum states cannot be perfectly copied. We investigate the precise limits on quantum cloning, the most general method of copying quantum information. We demonstrate that for certain quantum ensembles it is preferable to copy without using a "measure and replicate" strategy. We demonstrate further that the information in the classically reducible ensemble is entirely copyable, while the information in the limiting case of an ensemble with identical states (dubbed the "purely quantum ensemble") is strictly conserved. Finally, we propose a physical model of optimal cloning, involving passing off-resonance excited atoms through optical microcavities.

An Erbium-Ytterbium Doped Fiber Laser

Todd Stievater

We report the design and construction of an erbium-ytterbium doped all fiber laser, capable of lasing at a wavelength of 1.56 um using a 1.06 um Nd:YAG as a pump source. Continuous wave (cw) lasing occurs at pump powers greater than about 50 mW with a slope efficiency of 1.6%. Pulsed lasing is attained with the addition of a nonlinear optical loop mirror (NOLM), producing stable pulse trains at the fundamental repetition rate of the resonator, 1.74 MHz. Pulses having a minimum duration of about 100 ns with a significant cw background have been achieved with feedback from a moving mirror, implying that mode-locked operation of the laser is not yet occurring.

PSYCHOLOGY

Children's Understandings of Family Arguments: Development in the Family Context

Paul A. Boxer

We investigated how children's reasoning about family arguments changed across the childhood age span, and how their reasoning was related to their behavior patterns. Ninety-two children, aged 5 through 12, from 60 families from the Berkshire area in northwestern Massachusetts, responded to questions about what caused and who could solve family arguments. Children listened either to a simulated parent-parent or parent-child argument and responded to one of our two argument scales. Children in each test were divided into three age groups for analysis. Their reasoning about family arguments was more similar over age groups than we had hypothesized, as demonstrated by their responses to subscales from and rank-ordering of items from the argument scales. Younger children (ages 5-6) were more likely than middle (ages 7-9) and older children (ages 10-12) to think they could solve marital arguments; they were also more likely to fault one parent more than the other as a cause for marital arguments. Across ages, children were hesitant to blame marital arguments on negative and enduring traits of the mother. Generally, children attributed marital and parent-child arguments more to systemic causes (causes involving both parties to an argument) than to causes involving one actor or the other. Exploratory hypotheses investigating the relationship between attributions and problem behavior were not confirmed; there was no relationship between the types of attributions children endorsed strongly and the types of behavior patterns they displayed based on a parent-report measure. Future studies should consider children's perspectives on family interactions as well as focusing on more complex and subtle ways that children's developmental levels and behavioral patterns might relate to their reasoning about family arguments.

Contextual Factors in Men's and Women's Self-Presentation of Academic Achievement

Laura B. Brown

Both men's and women's self-presentation of academic achievement were investigated. Subjects anticipated interaction with a confederate, who was either self-promoting, moderate, or self-deprecating discussing his or her academic achievement. Subjects then predicted their own first semester GPA, believing they would subsequently discuss this with the confederate. Predicted GPAs were adjusted with actual GPAs to control for any effects of actual academic performance. Men's, but not women's, (adjusted) predicted GPAs were affected by the confederate's behavior. Men predicted the highest in the self-promoting condition, and, in fact over, predicted their GPAs in that condition. They also predicted the lowest in the self-deprecating condition.

The subjects' perceptions and affect were also examined. The self-promoting confederate was liked the least and by both men and women. Female participants rated the self-promoting confederate as the least smart, while men's ratings of the confederate did not vary based on the confederate's behavior. Similarly, female confederates who were self-deprecating were rated by participants as the least smart, while ratings of a male confederate's intelligence did not vary based on his self-presentation of achievement. Also, based on self-report measures, women tended to feel more comfortable with the self-deprecating confederate whereas men tended to feel more comfortable with the self-promoting confederate. Women were more uncomfortable than men predicting their GPAs. An analysis of qualitative statements of modesty or immodesty did not yield any significant results. The findings are reviewed in light of self-in-relation theory and the variable, self-serving, and contextual nature of self-presentation.

Pretense Comprehension in Young Children: Effects of Representation and Social Context

Amelia Cottrell

Pretense plays an important role in the cognitive-social development of young children. Recent research has demonstrated that by age two children engage in shared pretend play with both caregivers and familiar peers. One question that has arisen from this research is what the young child comprehends when a play partner enacts a make-believe scenario. For example, do young children truly understand the pretend nature of the interactive games they play (e.g., during a make-believe tea party), or do they rely on cues and prompts to comprehend their partner's non-literal actions? To answer this question, researchers have designed studies to determine experimentally at what age children begin to understand pretense.

This thesis consisted of two studies that addressed several factors that may influence children's comprehension of pretend actions. In Experiment 1, twelve older one-year olds (M = 21 months) and twelve younger two-year olds (M = 30 months) were tested for their comprehension of pretend actions performed on an animal (e.g., pretending to "pour" milk on a giraffe's neck). After each trial, children were presented with a choice array consisting of either pictures or toys that showed the animal correctly transformed according to the pretend action (e.g. the giraffe with milk on its neck), the animal unchanged, and an incorrectly transformed animal (e.g., the giraffe with a red leg). Older children were correct on a majority of the trials, whereas younger children's answers did not differ from chance. The hypothesis of a difference between children's trials when choosing among pictures and among objects was not supported for either age group. Therefore, the failure of the younger children was not because of a representational problem with the choice array.

Experiment 2 addressed the issue of testing children within a supportive context. It is possible that the younger children failed in Experiment 1, not because they did not understand the pretense, but because they were overwhelmed by the unfamiliarity of the context within which the test took place. In Experiment 2, ten older one-year olds (M = 20 months) were tested in two sessions using the toy animals and pretend actions from Experiment 1. Now, however, the context was varied so that in one session children were tested by the experimenter and in one session they were tested by their mothers. No difference was found in the number of correct responses between the sessions and the children did not score above chance in either session, indicating that being tested by the mother did not help the children to understand the pretense.

The implications of these two studies are discussed. The conclusion that children below the age of two years are simply unable to comprehend someone else's pretense is presented along with alternative explanations for their failure. The absence of the effect of the mother as tester is considered as part of a larger discussion about the origins of pretense comprehension and symbolic thought more generally, with particular attention to the role of the naturally occurring interactions between caregivers and their young children.

The Effects of Computer Animation on Jury Decision Making

Meghan Dunn

This research examined the effects -- both facilitative and prejudicial -- of computer-animated displays on jury-decision making. Experiment 1 tested the accuracy hypothesis that verdicts would change in the direction implied by the physical evidence more with an animated presentation of the evidence than they would with only a verbal presentation. A mock trial paradigm was used and two versions of a case were created: in one the physical evidence supported the plaintiff; in the other it supported the defendant. The results supported the accuracy hypothesis. The differences in verdicts between the plaintiff and defense versions of the case were more pronounced when the case was accompanied by a computer-animated display than in the no-display control group.

Experiment 2 tested the prejudice hypothesis that computer-animated displays can bias the jury in favor of the side presenting the display. Using the same case as used in Experiment 1, some subjects received oral presentations of the plaintiff and defense theories, while others also saw the plaintiff or defense theory depicted in an animated display. Although the results were not quite significant, they were in the predicted direction. When subjects were presented with an animated presentation of the plaintiff's theory, verdicts shifted in favor of the plaintiff's theory, even when the physical evidence did not support that theory.

The results of Experiments 1 and 2 demonstrate that computer-animated displays can have a persuasive effect on jurors' verdicts. However, the nature of that effect depends on the display itself. If the animated display accurately represents the physical facts of an event, it can improve judgment accuracy and facilitate decision-making. If however, the animated display depicts theory and speculation, it can bias the jury in favor of the side presenting the display. To some extent, both accuracy and prejudice effects were demonstrated in the studies conducted.

Self-Group Discrepancy and Self-Handicapping Behaviors

Kristen Harrington

An experiment was conducted to explore the self-presentational strategies individuals employ after they are informed that they have deviated from an in-group norm. The Bem Sex Role Inventory (BSRI) was administered to assess the extent to which individuals identified with their gender group. On the basis of their responses to the BSRI, participants were categorized as either high or low sex-typed. Participants played a video game -Tetris - and were informed that they performed above, below or consistent with what was expected for their group. Participants then were allowed to choose to listen to either facilitating or inhibiting music while playing a subsequent video game with either an in-group or out-group member. Results revealed that although high sex-typed individuals did not experience higher levels of guilt and anxiety after being informed that their individual performance was discrepant from what was expected for their group, performance feedback did influence their tape choice. Whereas, high sex typed individuals given below average feedback chose tapes that enhance performance, those who received above average feedback chose debilitating music. Measures designed to uncover the motives for participants tape choice did not reveal many significant differences. The results are discussed in terms of the implications of self-categorization for peoples use of impression management strategies.

Screening for Dementia: A Brief Cognitive Screening Battery Sensitive to Alzheimer's Disease

Bridget Kelly

The goal of this research was to develop a brief cognitive screening battery that is sensitive to Alzheimer's Disease (AD) using a group of AD patients and healthy age- and education-matched controls. There is a need in the primary care setting for a screening battery sensitive to AD that is simple and takes less than ten minutes to administer. This study evaluated a battery which is short, requires only simple administration procedures and materials, and includes measures of memory, language, orientation, and visuospatial ability, the four areas of cognition which are impaired in AD. The subjects who were administered the screening battery were either AD patients (N=52) who were seen as part of their regular care at the Memory Disorders Clinic at Southwestern Vermont Medical Center in Bennington, VT or healthy elderly controls (n=52) who were either spouses of patients at the Memory Disorders Clinic or community residing elderly. Logistic regression analysis was used to derive a probability formula that would best discriminate between AD patients and controls. The derived formula classified the sample used in the logistic regression with a sensitivity of 97% and a specificity of 97%, and a cross-validation sample with a sensitivity of 94% and a specificity of 88%. The battery had an overall sensitivity of 96% and a specificity of 94%, misclassifying 2 of 52 AD patients and 3 of 52 healthy controls. The battery evaluated here shows promise as an accurate screening instrument sensitive to AD. The diagnostic accuracy of the battery and its sensitivity to AD and dementia could be further assessed by examining its performance both in patients with other primary degenerative disorders and in patients with depression, as well as in a study using undiagnosed patients in a primary care setting who are followed in a full diagnostic evaluation.

The Long-Term Implications of Prenatal Alcohol Exposure on
Thermoregulatory Adaptations in the Rat

Jennifer C. Kerns

Prenatal alcohol exposure has been shown to have short-term effects on the development of thermoregulatory mechanisms in rat pups. The present research served to examine the long-term effects of prenatal exposure to alcohol on thermoregulatory mechanisms in adult offspring. Pregnant dams were given either a liquid diet with 35% of the calories derived from ethanol, a liquid diet with 0% of the calories derived from ethanol to control for any nutritional effects of alcohol liquid diet administration, or a lab chow and water control diet. Adult offspring were either administered a cafeteria diet, which causes hyperphagia and results in increased brown adipose tissue (BAT) activity in normal rats; were acclimated to a 4[[ring]]C environment, which causes similar increased BAT activity in normal rats; or were left unchallenged. Subjects were weighted regularly throughout the postnatal challenge period. Interscapular BAT was excised, and NE content was analyzed through high performance liquid chromatography (HPLC). As expected, BAT size increased in response to both the cafeteria diet and chronic exposure to cold. Pair-fed animals displayed significantly lower increases in BAT weight in response to cold-acclimation. In addition, pair-fed animals demonstrated significantly lower BAT concentrations of NE in the cafeteria diet condition. Such pair-fed effects may be a result of prenatal stress rather than maternal malnutrition. In general, animals exposed to cold environments had larger BAT deposits and lower concentrations of NE than did cafeteria diet-fed or unchallenged animals. Such decreases in NE concentration in the BAT may be a result of increased BAT volume combined with an unchanged level of sympathetic nervous system innervation. The teratogenic effects of alcohol could not be separated from nutritional effects, and appeared to have no significant long-term implications on thermoregulation in the rat.

Longitudinal Change in Probable Alzheimer's Disease As Measured by a
Brief Cognitive Screening Battery

Mahri Relin

In 1994, Aliina Hirshoff developed a screening battery for Alzheimer's disease to be used by primary care physicians that was designed to be brief, inexpensive, and easy to administer. The battery has been shown to be extremely sensitive to primary degenerative dementia (Hirschoff, 1994; Kelly, 1995), however its rate of change during disease progression has not yet been tested. The purpose of this study was to determine the sensitivity of the battery to the progressive decline in functioning over time in Alzheimer's patients. Subjects were patients (n=17) and spouses of patients (n=13) who visited the Memory Disorders Clinic in Bennington, Vermont. All subjects were given the Hirschoff battery twice -- once during their initial visit to the clinic and once during a follow-up visit that occurred between 6 and 30 months later. Rates of score change on the test were correlated whenever possible with two well-known rating instruments: The Mini-Mental Status Examination (MMSE) and the Alzheimer's Disease Assessment Scale (ADAS). A probability of having probable Alzheimer's disease (PRAD, 0-100 percent) was determined for each subject according to their scores on the battery. Among the PRAD patients, the average monthly rate of change in probability was 0.18 percent. The average monthly rate of change for controls was -0.68 percent. Rates of change did not correlate well with change on the ADAS or the MMSE. Among the individual subtests, Enhanced Cued Recall and Category Fluency were the only measures on which the PRADs significantly differed from controls in score change per month. Based on these results, the Hirschoff battery does not seem to be a sensitive measure of change during the course of the Alzheimer's disease. It is possible that such a short battery cannot be used as both a diagnostic and progression tool.

Children's Attributions About Family Conflict: The Call of Children's "Stories"

Heather Weston

We investigated children's reasoning about family arguments across the childhood age span. One of the goals of this thesis was and is to aid clinicians, family therapists especially, in incorporating children's views of family problems into their theories and interventions with families. Reviews of studies investigating children's perceptions of family conflict and their socio-cognitive development are presented. In the present study, ninety-two children, aged 5 to 12, from 60 Berkshire County families listened either to a simulated parent-parent (Test 1) or parent-child (Test 2) argument and responded to one of our two scales about what caused and who could solve family arguments. For analysis, children in each test were divided into three age groups: 5 to 6 year-olds, 7 to 9 year-olds, and 10 to 12 year-olds. Their reasoning about family arguments was more similar over age groups than we hypothesized, as demonstrated by their responses to our subscales and by rank-orderings of individual scale items. Regarding marital arguments, younger children were more likely than older children to think they could solve these arguments; they were also more likely to fault one parent more than the other. Over all ages, children were very hesitant to blame marital arguments on negative and enduring traits of the mother. Exploratory findings linking children's reasoning family environment (Moos, 1974) revealed a higher level of family conflict correlated with increased mother-blaming and child- blaming attributions. Future studies and clinical interventions should consider children's perspectives on family interactions, as well as more complex and subtle ways that development and family environment might relate to children's reasoning about family arguments.

Postnatal Cocaine Exposure's Effects on
Neurochemical Mechanisms of Juvenile Play Behavior in Rats

Stacy Whitcomb

Every year about 100,000 cocaine exposed children are born, and researchers continue to debate over the behavioral implications of this exposure. In attempts to settle this debate, animal models have been used, since human studies are confounded by polydrug use populations and inaccuracy of self reporting measures. This study served to investigate the effects postnatal cocaine exposure elicited on juvenile play behavior and the areas of the brain believed to be involved in the generation and regulation of play behavior. The postnatal paradigm mimics third trimester exposure in humans. A sex reversal of play behavior was expected, in that cocaine would elicit less active play behavior (pins, bites, mounts) and more passive behavior (getting pinned, mounted, bitten, non-play contact) in male subjects and females would show more active and less passive play than controls. It was also expected that neurochemical deviations would be seen in all the brain structures investigated. Seventy-two Long-Evans hooded rats served as subjects (Ss) in one of four conditions, procedural controls (PC), No Injections (NI), Saline Injections (SI), or Cocaine Exposed (CE) at 20 mg/kg ip injections. At postnatal day (PND) 31, Ss were observed for quantifiable play behaviors (pinning, getting pinned, mounting, getting mounted, biting, defending bites, boxing, non-play contact, burrowing, self-grooming, and rearing) while playing with a stimulus pup. Levels of norepinephrine (NE), dopamine (DA) and its metabolite, DOPAC were measured in the pre-frontal cortex (PFC), striatum, hypothalamus, and amygdala. Cocaine pups weighed less than controls on PND 4-12, but caught up by PND 31. Significant effects across condition were not seen from play behaviors, but PC pups had higher activity levels than controls and also had higher levels of NE in the PFC. A significant interaction was seen for DA levels in the striatum in that CE males had less DA than controls and CE females had more DA than controls. Data suggest effects seen may be due to the stress of handling procedures, and that cocaine may be a very mild teratogen that affects growth short-term. It also appears to affect the striatal dopaminergic system, however, the resultant behavioral effects were beyond the scope of this study and should be investigated further.

 


Modified by: bbabcock
Modification Date: December 13, 1995