Bovine adenovirus (BAdV)-3 is a non-enveloped, icosahedral virus with a double-stranded DNA genome, and is being developed as a vector for vaccination of animals and humans. Expression of viral genes is divided into early, intermediate, and late phases. The late genes of BAdV-3 are grouped into seven families (L1 to L7) based on usage of common polyadenylation site(s). The L1 region of BAdV-3 encodes the 52K protein, a non-structural protein conserved among members of the family Adenoviridae. In human adenovirus (HAdV)-5, the 52K protein is involved in packaging of the viral DNA into the capsid. The N-terminal half of the protein has been proposed to mediate serotype specificity of DNA packaging. The objective of this study was to characterize the 52K protein of BAdV-3. <p> DNA sequence analysis revealed that the BAdV-3 52K open reading frame encodes a protein of 370 amino acids rather than 331 amino acids as previously reported. Western blotting with anti-52K serum detected the expression of a 40kDa protein at 24 to 72 hrs post-infection. BAdV-3 52K localized predominantly to the nucleus in BAdV-3 infected cells and in transfected cells in the absence of other viral proteins. Analysis of mutant 52K proteins revealed that residues 102-110 were necessary but not sufficient for nuclear import. This suggests that residues upstream or downstream of the identified 52K nuclear localization signal (NLS) are required, or that the function of the NLS is dependent on its conformation within 52K. <p> The nuclear import of 52K is significantly, but not completely, dependent on soluble factors, ATP, and temperature. A peptide competing for binding to importin beta and a peptide encoding the NLS of Ycbp80 were also able to inhibit nuclear import of 52K. However, a dominant negative mutant of Ran was unable to block 52K nuclear import. These results suggest that 52K uses a classical importin alpha/importin beta pathway for nuclear import. In support of this, a specific interaction between 52K and importin alpha-3 was detected. In addition, 52K was able to accumulate in the nucleus in the absence of soluble factors and ATP when the nuclear membrane was permeabilized with detergent. This suggests that, in addition to nuclear import by the importin alpha/importin beta pathway, 52K is able to accumulate in the nucleus by binding to nuclear components. <p> A yeast two-hybrid system identified interactions between BAdV-3 52K and pV, pVI, pVII, and IVa2. However, only the interaction with pVII could be confirmed by GST pulldown. 52K and pVII also interact during BAdV-3 infection. An interaction between 52K and pVII has previously been shown in HAdV-5 infected cells. <p> Mass spectrometry analysis of proteins co-precipitating with BAdV-3 52K identified a cellular protein, NFkB-binding protein (NFBP), which interacted with 52K. The interaction between NFBP and 52K was confirmed <i>in vitro</i> and <i>in vivo</i>. NFBP has been shown to be essential for ribosomal RNA (rRNA) processing. While NFBP is normally localized in the nucleolus, co-expression with 52K results in the redistribution of NFBP from the nucleolus to other parts of the nucleus. While this suggested that redistribution of NFBP by 52K could inhibit rRNA processing during BAdV-3 infection, we were unable to detect a difference in rRNA processing in cells expressing truncated or full-length 52K in the absence of other viral proteins. Since NFBP is a multi-functional protein, future experiments should focus on other possible biological functions of the interaction of NFBP with BAdV-3 52K.

<p>RPL23a is one of the ~80 ribosomal proteins (r-proteins) of the cytoplasmic ribosome in the model plant <i>Arabidopsis thaliana</i>. The objectives of this research were to establish Arabidopsis RPL23a as a functional r-protein, characterize expression patterns for the two genes (RPL23aA and B) encoding RPL23a using reverse transcription PCR (RT-PCR), and identify regulatory elements controlling the expression of RPL23aA and B. Complementation of a yeast l25 mutant with AtRPL23aA demonstrated that AtRPL23aA can fulfill all the essential functions of L25 in vivo. A survey of various Arabidopsis tissue types showed that, while RPL23aA and B expression patterns both showed increased transcript abundance in mitotically active tissues, RPL23aB transcript levels were generally lower than those of RPL23aA and responded differently to abiotic stresses. In order to determine cis regulatory elements controlling RPL23aA and B expression, the 5 regulatory region (RR) of each gene was characterized via plants carrying a series of 5 RR deletion fragments upstream of a reporter. Transcript start sites and 5 untranslated regions (UTRs) for both RPL23aA and B were also characterized using primer extension, and transcripts from 5 deletion transgenics were amplified using RT-PCR. No correlation was observed between putative cis-acting elements and the expression patterns from the RPL23aA and B deletion transgenics, although a 102 bp sequence in the RPL23aB 5 RR was found to contain pollen-specific elements. 5 leader introns were found in each RPL23a gene, and amplification of transgene transcripts from deletion series plants indicated the importance of post-transcriptional and translational regulation in RPL23aA and B expression. This thesis work is the first demonstration of a plant RPL23a protein as a functional member of the L23/L25 (L23p) conserved r-protein family, and is one of the few in-depth studies of the regulation of r-protein genes in plants. While the majority of previous research on plant r-protein gene expression has focused solely on transcript levels, I show herein that post-transcriptional mechanisms have a critical role in regulating these genes, and thus plant r-protein genes more strongly resemble their mammalian counterparts than those of yeast in terms of structure and regulation.

Ribosome-inactivating proteins (RIPs) such as Shiga-like toxin 1 (SLT-1) halt protein synthesis in eukaryotic cells by depurinating a single adenine base in the sarcin-ricin loop of 28S rRNA. The molecular details involved in the ER lumenal escape and subsequent site-specific depurination are lacking, despite a general understanding of the biochemical basis of SLT-1 toxicity. Using a combination of yeast-2-hybrid and HeLa lysate pull-down followed by LC-MS/MS we have discovered yeast and human proteins that interact with the catalytic A1 chain of SLT-1. Yeast-2-hybrid library screens followed by the expression of full-length protein candidates and pull-down experiments yielded Cue2 as the only yeast cellular component that binds to the SLT-1 A1 chain. Further truncational analysis revealed that the known protein domains (two Cue domains and a Smr domain) within the primary sequence of Cue 2 were not essential for the interaction. Cue2 is a yeast monoubiquitin binding protein of no known function that is structurally homologous to the human ubiquitin-associated domain which has been implicated in intracellular routing and ER-associated degradation. Pull-down experiments indicated that the mechanism by which the catalytic domain of RIPs cleaves its substrate involves initial docking interactions with the ribosomal stalk by virtue of a conserved acidic C-terminal peptide domain common to all three stalk proteins P0, P1, and P2. The A1 chain of SLT-1 transiently binds to this peptide with a modest binding constant and rapid on and off rates. Mutagenesis of charged residues within the A1 chain identified a cationic surface that interacts with the peptide motif. In addition, phage-display was used to rapidly probe the importance of each residue within this C-terminal ribosomal peptide. The analysis revealed a complementary acidic surface and an additional hydrophobic motif involved in the interaction. Moreover, deletion mutagenesis performed on the ribosomal protein P0 revealed that the A1 chain binds to an alternate site on P0 in proximity to the contact sites for P1/P2 heterodimers. These results demonstrate that the catalytic chain of RIPs such as SLT-1 dock on ribosomes using two classes of binding sites located within the ribosomal stalk which may aid in orienting their catalytic domain in close proximity to the depurination site.

Bovine adenovirus (BAdV)-3 is a non-enveloped, icosahedral virus with a double-stranded DNA genome, and is being developed as a vector for vaccination of animals and humans. Expression of viral genes is divided into early, intermediate, and late phases. The late genes of BAdV-3 are grouped into seven families (L1 to L7) based on usage of common polyadenylation site(s). The L1 region of BAdV-3 encodes the 52K protein, a non-structural protein conserved among members of the family Adenoviridae. In human adenovirus (HAdV)-5, the 52K protein is involved in packaging of the viral DNA into the capsid. The N-terminal half of the protein has been proposed to mediate serotype specificity of DNA packaging. The objective of this study was to characterize the 52K protein of BAdV-3. <p> DNA sequence analysis revealed that the BAdV-3 52K open reading frame encodes a protein of 370 amino acids rather than 331 amino acids as previously reported. Western blotting with anti-52K serum detected the expression of a 40kDa protein at 24 to 72 hrs post-infection. BAdV-3 52K localized predominantly to the nucleus in BAdV-3 infected cells and in transfected cells in the absence of other viral proteins. Analysis of mutant 52K proteins revealed that residues 102-110 were necessary but not sufficient for nuclear import. This suggests that residues upstream or downstream of the identified 52K nuclear localization signal (NLS) are required, or that the function of the NLS is dependent on its conformation within 52K. <p> The nuclear import of 52K is significantly, but not completely, dependent on soluble factors, ATP, and temperature. A peptide competing for binding to importin beta and a peptide encoding the NLS of Ycbp80 were also able to inhibit nuclear import of 52K. However, a dominant negative mutant of Ran was unable to block 52K nuclear import. These results suggest that 52K uses a classical importin alpha/importin beta pathway for nuclear import. In support of this, a specific interaction between 52K and importin alpha-3 was detected. In addition, 52K was able to accumulate in the nucleus in the absence of soluble factors and ATP when the nuclear membrane was permeabilized with detergent. This suggests that, in addition to nuclear import by the importin alpha/importin beta pathway, 52K is able to accumulate in the nucleus by binding to nuclear components. <p> A yeast two-hybrid system identified interactions between BAdV-3 52K and pV, pVI, pVII, and IVa2. However, only the interaction with pVII could be confirmed by GST pulldown. 52K and pVII also interact during BAdV-3 infection. An interaction between 52K and pVII has previously been shown in HAdV-5 infected cells. <p> Mass spectrometry analysis of proteins co-precipitating with BAdV-3 52K identified a cellular protein, NFkB-binding protein (NFBP), which interacted with 52K. The interaction between NFBP and 52K was confirmed <i>in vitro</i> and <i>in vivo</i>. NFBP has been shown to be essential for ribosomal RNA (rRNA) processing. While NFBP is normally localized in the nucleolus, co-expression with 52K results in the redistribution of NFBP from the nucleolus to other parts of the nucleus. While this suggested that redistribution of NFBP by 52K could inhibit rRNA processing during BAdV-3 infection, we were unable to detect a difference in rRNA processing in cells expressing truncated or full-length 52K in the absence of other viral proteins. Since NFBP is a multi-functional protein, future experiments should focus on other possible biological functions of the interaction of NFBP with BAdV-3 52K.

Ribosome-inactivating proteins (RIPs) such as Shiga-like toxin 1 (SLT-1) halt protein synthesis in eukaryotic cells by depurinating a single adenine base in the sarcin-ricin loop of 28S rRNA. The molecular details involved in the ER lumenal escape and subsequent site-specific depurination are lacking, despite a general understanding of the biochemical basis of SLT-1 toxicity. Using a combination of yeast-2-hybrid and HeLa lysate pull-down followed by LC-MS/MS we have discovered yeast and human proteins that interact with the catalytic A1 chain of SLT-1. Yeast-2-hybrid library screens followed by the expression of full-length protein candidates and pull-down experiments yielded Cue2 as the only yeast cellular component that binds to the SLT-1 A1 chain. Further truncational analysis revealed that the known protein domains (two Cue domains and a Smr domain) within the primary sequence of Cue 2 were not essential for the interaction. Cue2 is a yeast monoubiquitin binding protein of no known function that is structurally homologous to the human ubiquitin-associated domain which has been implicated in intracellular routing and ER-associated degradation. Pull-down experiments indicated that the mechanism by which the catalytic domain of RIPs cleaves its substrate involves initial docking interactions with the ribosomal stalk by virtue of a conserved acidic C-terminal peptide domain common to all three stalk proteins P0, P1, and P2. The A1 chain of SLT-1 transiently binds to this peptide with a modest binding constant and rapid on and off rates. Mutagenesis of charged residues within the A1 chain identified a cationic surface that interacts with the peptide motif. In addition, phage-display was used to rapidly probe the importance of each residue within this C-terminal ribosomal peptide. The analysis revealed a complementary acidic surface and an additional hydrophobic motif involved in the interaction. Moreover, deletion mutagenesis performed on the ribosomal protein P0 revealed that the A1 chain binds to an alternate site on P0 in proximity to the contact sites for P1/P2 heterodimers. These results demonstrate that the catalytic chain of RIPs such as SLT-1 dock on ribosomes using two classes of binding sites located within the ribosomal stalk which may aid in orienting their catalytic domain in close proximity to the depurination site.

<p>RPL23a is one of the ~80 ribosomal proteins (r-proteins) of the cytoplasmic ribosome in the model plant <i>Arabidopsis thaliana</i>. The objectives of this research were to establish Arabidopsis RPL23a as a functional r-protein, characterize expression patterns for the two genes (RPL23aA and B) encoding RPL23a using reverse transcription PCR (RT-PCR), and identify regulatory elements controlling the expression of RPL23aA and B. Complementation of a yeast l25 mutant with AtRPL23aA demonstrated that AtRPL23aA can fulfill all the essential functions of L25 in vivo. A survey of various Arabidopsis tissue types showed that, while RPL23aA and B expression patterns both showed increased transcript abundance in mitotically active tissues, RPL23aB transcript levels were generally lower than those of RPL23aA and responded differently to abiotic stresses. In order to determine cis regulatory elements controlling RPL23aA and B expression, the 5 regulatory region (RR) of each gene was characterized via plants carrying a series of 5 RR deletion fragments upstream of a reporter. Transcript start sites and 5 untranslated regions (UTRs) for both RPL23aA and B were also characterized using primer extension, and transcripts from 5 deletion transgenics were amplified using RT-PCR. No correlation was observed between putative cis-acting elements and the expression patterns from the RPL23aA and B deletion transgenics, although a 102 bp sequence in the RPL23aB 5 RR was found to contain pollen-specific elements. 5 leader introns were found in each RPL23a gene, and amplification of transgene transcripts from deletion series plants indicated the importance of post-transcriptional and translational regulation in RPL23aA and B expression. This thesis work is the first demonstration of a plant RPL23a protein as a functional member of the L23/L25 (L23p) conserved r-protein family, and is one of the few in-depth studies of the regulation of r-protein genes in plants. While the majority of previous research on plant r-protein gene expression has focused solely on transcript levels, I show herein that post-transcriptional mechanisms have a critical role in regulating these genes, and thus plant r-protein genes more strongly resemble their mammalian counterparts than those of yeast in terms of structure and regulation.

Eukaryotic cytosolic ribosomes are composed of two distinct subunits consisting of four individual ribosomal RNAs and, in Arabidopsis thaliana, 81 ribosomal proteins. Functional subunit assembly is dependent on the production of each ribosomal component. Arabidopsis thaliana r-protein genes exist in multi-gene families ranging in size from two to seven transcriptionally active members. The cytosolic RPS15a gene family consists of four members (RPS15aA, -C, -D and -F) that, at the amino acid level, share 87-100% identity. Using semi-quantitative RT-PCR I have shown that RPS15aC is not expressed and that transcript abundance differs both spatially and temporally among the remaining RPS15a genes in non-treated Arabidopsis tissues and in seedlings following a variety of abiotic stresses. A comprehensive analysis of the RPS15a 5' regulatory regions (RRs) using a series of deletion constructs was used to determine the minimal region required for gene expression and identify putative cis-regulatory elements. Transcription start site mapping using 5' RACE indicated multiple sites of initiation for RPS15aA and -F and only a single site for RPS15aD while all three genes contain a leader intron upstream of the start codon. Analysis of reporter gene activity in transgenic Arabidopsis containing a series of 5' RR deletion::GUS fusions showed that, similar to previous RT-PCR results, there was a trend for mitotically active tissues to stain for GUS activity. Putative cis-elements including the TELO box, PCNA Site II motif and pollen specific elements were identified. However, there was not always a clear correlation between the presence of a putative element and RPS15a transcript abundance or GUS activity. Although variation in transcriptional activity of each RPS15a gene has been observed, subcellular localization of both RPS15aA and -D in the nucleolus has been confirmed in planta by confocal microscopy. The results of this thesis research suggest while all three active RPS15a genes are transcriptionally regulated, additional post-transcriptional and/or translational regulation may be responsible for final RPS15a levels while differential isoform incorporation into ribosomal subunits may be the final point of r-protein regulation.

Ribosomal proteins (RPs), synthesized in the cytoplasm, need to be transported from the cytoplasm to the nucleolus (a nuclear compartment), where a single molecule of each RP assembles with rRNAs to form the large and small ribosomal subunits. The objectives of this research were to identify nuclear/nucleolar localization signals (NLSs/NoLSs; generally basic motifs) that mediate the transport of Arabidopsis RPL23aA, RPL15A and RPS8A into the nucleus and nucleolus, and to study transcriptional regulation and subcellular localization of RPs. While all previous research has shown that nucleolar localization of proteins is mediated by specific basic motifs, in this study, I showed that a specific number of basic motifs mediated nucleolar localization of RPL23aA, rather than any specific motifs. In this protein, single mutations of any of its eight putative NLSs (pNLSs) had no effect on nucleolar localization, however, the simultaneous mutation of all eight completely disrupted nucleolar localization, but had no effect on nuclear localization. Furthermore, mutation of any four of these pNLSs had no effect on localization, while mutation of more than four increasingly disrupted nucleolar localization, suggesting that any combination of four of the eight pNLSs is able to mediate nucleolar localization. These results support a charge-based system for the nucleolar localization of RPL23aA. While none of the eight pNLSs of RPL23aA were required for nuclear localization, in RPS8A and RPL15A, of the 10 pNLSs in each, the N-terminal two and three NLSs, respectively, were absolutely required for nuclear/nucleolar localization. Considering the presence of only a single molecule of each RP in any given ribosome, which obligates the presence of each RP in the nucleolus in equal quantities, I studied transcriptional regulation of Arabidopsis RP genes and the subcellular localization of five RP families to determine the extent of coordinated regulation of these processes. Variation of up to 300-fold was observed in the expression levels of RP genes. However, this variation was drastically reduced when the expression level was considered at the RP gene family level, indicating that coordinate regulation of expression of RP genes, coding for individual RP isoforms, is more stringent at the family level. Subcellular localization also showed differential targeting of RPs to the cytoplasm, nucleus and nucleolus, together with a significant difference in the nucleolar import rates of RPS8A and RPL15A. Although one could expect coordinated regulation of the processes preceding ribosomal subunit assembly in the nucleolus, my results suggest differential regulation of these processes.

Babesia odocoilei is an intraerythrocytic apicomplexan parasite which infects cervidae, sometimes causing babesiosis. It is vectored by the tick Ixodes scapularis and is distributed throughout the southeastern United States. The geographic and host range continue to extend as new incidence of infection is detected. A genomic DNA region spanning the internal transcribed spacer 1 (ITS1), 5.8S rRNA gene, and ITS2 of ribosomal RNA (rRNA) from 18 B. odocoilei isolates (speciation confirmed by small subunit rRNA analysis) was amplified using the polymerase chain reaction, cloned and sequenced. The isolates originated from 6 different cervidae or bovidae hosts in various U.S. geographic areas. Included in the analysis was a previously described reindeer B. odocoilei-like isolate, RD61, which showed only 99.0% identity in SSU rRNA analysis to B. odocoilei. Percent identity pairwise comparisons among the samples were calculated for both the full ITS1-5.8S-ITS2 and individual genomic regions. Identity values for all comparisons ranged from 90% to 100%, with the exception of RD61, which showed no higher than 88% identity for all gene regions. An analysis of fixed differences identified in the ITS1 and ITS2 gene regions of all clones revealed 21 fixed differences in ITS1, and only 11 in ITS2. Most isolates were found to have 2 overall patterns of fixed differences, although some had 1 or 3. Phylogenetic analysis of all sequences for the entire ITS1-5.8S-ITS2 gene region placed most isolates into 2 distinct groups corresponding to those observed in the analysis of fixed differences. This suggested the presence of at least 2 rRNA transcription units in B. odocoilei. ITS analysis failed to demonstrate host or geographic differences that might serve to pinpoint the source of outbreaks of B. odocoilei in farmed and managed host animals. This failure might result from genetic recombination of ITS genomic regions during the tick vector stage. Lack of conspecificity between the RD61 isolate and B. odocoilei was supported by this study; however, more data are needed to clarify the taxonomic status of this B. odocoilei-like isolate.

Thesis (Ph. D.)--University of Washington, 2000. / Vita. Includes bibliographical references (leaves 70-81).