Orientador: João Pessoa Araújo Junior / Banca: Paulo Eduardo Brandão / Banca: Alexandre Secorum Borges / Resumo: A cinomose e a parvovirose canina são duas enfermidades infecto-contagiosas de grande importância para a clínica de pequenos animais onde a PCR vem sendo aplicada com ótimos resultados no diagnóstico. No entanto, para o sucesso da técnica, é necessário que o ácido nucléico esteja o mais puro possível e livre de inibidores das polimerases (Transcriptase reversa e/ou Taq DNA polimerase), desejando-se um método de extração simples e rápido. O teste de coaglutinação utilizando o Staphylococcus aureus (COA) é baseado na propriedade da proteína A de se ligar especificamente à porção Fc da imunoglobulina G de alguns mamíferos e algumas subclasses de IgG de camundongos. Assim, neste trabalho utilizou-se a coglutinação para obtenção de DNA ou RNA livres de inibidores, com capacidade de concentração de partículas virais dispersas nas amostras biológicas e de forma simples, rápida e de baixo custo. Para tanto, 10 amostras de fezes positivas para o vírus da parvovirose canina e 17 amostras de urina positivas para o vírus da cinomose foram submetidas à extração de ácidos nucléicos utilizando o COA e kits comerciais para posteriormente serem analisadas pela PCR em tempo real e PCR convencional respectivamente. As amostras de fezes foram diluídas de 1: 10 a 1: 100 000 e as amostras de urina foram utilizadas puras. A metodologia desenvolvida foi eficiente na extração dos dois tipos de amostra. O método proposto demonstrou ser confiável e de baixo custo para a preparação de DNA e RNA viral para o diagnóstico molecular. / Abstract: PCR presents excellent results for the diagnosis of canine distemper and canine parvoviruses, two important infectious and contagious diseases for small animal internal medicine. However, success of technique depends on nucleic acid samples free of polymerase inhibitors (Reverse Transcriptase and / or Taq DNA polymerase). The coagglutination test using Staphylococcus aureus (COA) is based on the property of specific binding of protein A to the Fc portion of immunoglobulin G of some mammals and some of IgG subclasses of mice. This work was carried out the coagglutination procedure to obtain nucleic acid inhibitors free, with capacity for viral particle concentration dispersed in biological samples, simply, quickly and low cost. For this purpose, 10 canine parvovirus positive stool samples and 17 canine distemper virus positive urine samples were submitted to the preparation of nucleic acids using the COA and commercial kits in order to be analyzed by real-time PCR or conventional PCR respectively. Fecal specimens were diluted from 1: 10 to 1: 100 000 and urine samples were used pure. The developed methodology was efficient in extracting the two types of sample. The method proposed demonstrated to be reliable and cheap to prepare viral DNA or RNA for molecular diagnosis. / Mestre

Six mouse monoclonal antibodies have been isolated which react against a bacterial fusion protein containing amino acids 364 to 623 of the NS-1 protein of the prototype strain of the Minute Virus of Mice (MVMp). All six were found to be of the IgG class of antibodies; five being IgG₁ and the sixth being IgG₂[formula omitted]. By immunoblot analyses, these antibodies all recognize an 83 kDa protein found only in MVM-infected mouse fibroblast cells, leading to the assumption that they are all NS-1 specific. Further evidence for this assumption is obtained from indirect immunofluorescence studies showing all but one of the mAbs react against a nuclear protein found in MVM-infected cells. The epitopes of the antibodies were mapped using carboxy-terminal deleted bacterial fusion proteins derived from the plasmid encoding the original antigen. For the six monoclonal antibodies, four distinct epitopes were found (A - D). Three were clustered in a 16 amino acid region near the carboxy-terminal of the bacterial fusion protein, while the fourth was slightly more toward the amino-terminal side. Competition ELISAs against a 25 amino acid NS-1 specific peptide confirmed the mapping of the A epitope recognized by the CE10 and AC6 monoclonal antibodies. Also in this thesis, the characterization of a NS-1 fusion protein and a non-fused NS-1 protein expressed in insect cells by recombinant baculoviruses is also described. The latter, a full-length NS-1 protein designated NS-1[formula omitted]ⅽ, was found to be an 84 kDa cytoplasmic protein. This protein was immunoprecipitated by all six monoclonal antibodies. A CE10 monoclonal antibody immunoaffinity column was employed in the single-step purification of NS-1 [formula omitted]c from insect cells. Four elution methods (alkaline, peptide, 6M guanidinium, and acid) were examined and the best purification was obtained using the acid elution. / Medicine, Faculty of / Biochemistry and Molecular Biology, Department of / Graduate

The positions of sequences necessary for transcription from the promoter located at map unit 4 of the bovine parvovirus (BPV) genome were determined. The autonomous parvoviruses, of which BPV is a member, contain two transcriptional units with promoters active in temporal order during infection. BPV proteins also appear in the same temporal order; the nonstructural (NS) proteins are produced before the capsid proteins. Northern blot analysis of BPV RNAs suggest that, like human parvovirus B19, all transcripts of BPV are initiated from a single promoter. A reporter construct was created by cloning the sequences from BPV containing the TAT A box located at nucleotides (nt) 250 - 254 upstream of the luciferase gene. A series of mutants were generated by deletion of restriction endonuclease fragments. Expression was assayed by transient expression of the constructs in transfected bovine fetal lung (BFL) cells, derived from the natural host. The data indicates that expression can be directed from the sequences containing the TAT A box. Analysis of expression of the deletion clones show that sequences from nt 120 - 170 of the BPV genome are also required for transcription. A search of these sequences reveals at least two consensus binding sites for cellular transcription factors. These are AP-I and the major late transcription factor (MLTF). MLTF has been shown to induce transcription from the early promoter of the dependovirus adena-associated virus (AA V). The presence of viral proteins provided in trans decreases expression from all constructs, with one exception. Expression, when nt 0 - 50 are deleted, is increased in the presence of mutant BPV NS-I. Feedback inhibition of P4 expression by NS-I, the P4 gene product, is seen in H-l, a rat parvovirus. The expression of BPV proteins in synchronized HeLa cells, which are not a natural tissue culture host was examined to determine the effect of a non-permissive host on BPV transcription. Expression of viral proteins in some parvoviruses is blocked in non-permissive cells. A block in transcription is seen during infection of non-permissive cells by B 19. AA V P5 expression is negatively regulated without co-infection of a helper virus, a non-permissive state. Luciferase activity was 570/0, compared to BFL cells. and a similar decrease in expression in the presence of viral proteins was observed. Viral nonstructural and capsid proteins could be detected by immunofluorescence, but only in the cytoplasm, suggesting that expression of viral proteins necessary for replication was not the block to a productive infection, but rather their translocation to the nucleus, as seen during restrictive H-l infection of transformed cells. BPV proteins have been observed localized to nuclear foci of transfected, synchronized BFL cells. The subcellular localization of viral proteins was detected by indirect immunofluorescence labeling using antiserum that recognizes both nonstructural and capsid proteins. Punctate nuclear staining has not been observed routinely during BPV infection or transfection of actively dividing cells. AA V proteins expressed in either BFL and HeLa cells, synchronized by hydroxyurea, were also observed in distinct nuclear foci. The same pattern of localization has been observed during co-infection of cells with AA V and adeno-virus, and during H-I infection. Gel mobility shift assays show that a cellular protein from BFL cells synchronized in S-phase recognizes and binds the right terminus of BPV. The right terminus, in hairpin and double-stranded linear forms, is an effective competitor, indicating the complex is specific, and suggesting that sequence, rather than structure may be the recognition signal for this cellular protein. The left terminus of BPV, in the hairpin conformation is also an efficient competitor for complex formation. It has been shown that a cellular protein forms a complex with the left terminus and that the right terminus is an efficient competitor for complex formation. This data taken together suggests the same cellular factor may recognize both termini, and correlates with the observation that both can act as origins of replication, and could be recognized by similar proteins. The heterologous terminus of AAV in the hairpin conformation, is not an efficient competitor for complex formation between the right terminus of BPV and a BFL cell protein. The lack of competition may support the indication that the recognition signal is a specific sequence rather than a particular secondary structure. / Ph. D.

Autonomous parvoviruses, such as bovine parvovirus (BPV), need a factor present at the S-phase of the cell cycle for a productive infection, while dependent parvoviruses, the adenoassociated viruses (AAVs), require a helper virus to complete an infectious cycle. However, AAV can replicate autonomously in synchronized cells, suggesting that an S-phase factor substitutes for the helper virus. To investigate the nature of the cellular S-phase factor, we performed DNA retardation assays with uninfected nuclear extracts of S-phase cells, synchronized by hydroxyurea pretreatment, and radiolabeled parvoviral termini in their hairpinned conformation. We observed that proteins in HeLa cells, a tissue culture host for AAV, specifically interacted with the terminal sequences of this virus, which act as origins of replication (oris). These assays also showed specific binding between S-phase cellular proteins and termini (oris) of heterologous parvoviruses, for which the cells are not a natural host. For example, proteins from bovine fetal lung (BFL) cells, a tissue culture host for BPV, were able to bind to an AAV terminus and HeLa cell proteins interacted with both termini of the BPV genome. All DNA-protein complexes investigated appeared to be specific for S-phase synchronized cells. In order to begin to characterize the protein(s) involved in the complex formation, we performed SDS-PAGE electrophoresis of some retarded complexes. We report that a 54 kd protein was contained in the complex formed with the BPV left terminus and BFL cell extract. [Binding of BFL cell proteins to a BPV left terminus has been reported earlier]. Using a similar technique, we observed that two phosphoproteins of 55 and 90 kd were present in the retarded complex formed between a BPV left terminus and HeLa cell extract. An antibody directed against human p53, an anti-oncoprotein, was shown to compete binding of BFL cell extract and HeLa cell extract to the BPV left terminus. This antibody also competed the binding of HeLa cell extract to the AAV terminus. Our data suggest that proteins with similar characteristics, most probably among which is p53, are involved in the ori-binding complexes, possibly exerting a role as positive regulator of parvoviral replication. The secondary structure of the viral ends is remarkably conserved among parvoviruses. Of particular interest is the presence of mismatched/unpaired nucleotides, forming a bubble, in the stem of the left hairpin of almost all autonomous parvoviruses. To analyze the possible role of these unpaired/mismatched nucleotides in the BPV life cycle, two mutants clones lacking the bubble region were constructed and their replicative properties were analyzed after electroporation in permissive cells. Infectivity of the mutant clones was determined by three techniques: observation of cytopathic effect, detection of virally-coded proteins by indirect immunofluorescence, and transient DNA replication assays. We report that the mutant clone containing duplicate sequences of the (mismatched) nucleotides numbered 46 to 57 (BLOP) was defective for replication. The other bubbleless clone (BLOM), that contains duplicate sequences of the (mismatched) nucleotides 99 to 105, was able to replicate. The later clone produced monomer-length viral DNA at about 20% of the level of the infectious genomic clone of BPV, when electroporated as a linear excised sequence. This clone was infectious since it could be propagated by subsequent passage. Expression of viral structural proteins was seen by an indirect immunofluorescence assay using anti-capsid antibodies. Our results suggest that the bubble in the left hairpin of BPV is not required for the viral life cycle, but that specific sequences within the mismatched/unpaired region are necessary for viral replication. / Ph. D.

Ph. D.

The polymerase activities involved in bovine parvoviral (BPV) DNA replication in vivo and in vitro have been described. In the in vitro system, purified and partially purified enzymes were used and the replication products were analyzed by gel electrophoresis and restriction enzyme digestion. DNA pol ૪ purified from fetal bovine liver, replicated BPV ssDNA to a unit-length covalently-linked dsDNA hairpin molecule but was unable to utilize purified BPV dsDNA as a template. Partially purified calf thymus DNA pol α replicated BPV DNA to a product 1 kbp smaller than unit-length dsDNA (11 kbp). Mapping of this product showed that the middle of the genome was under-represented. Purified pol a from bovine fetal lung (BFL) cells was capable of only end-labeling BPV ssDNA. If HeLa cell DNA pol α, which consisted of the core enzyme plus cofactors C₁ C₂, was used, the products consisted of both noncovalently- and covalently-linked unit-length dsDNA hairpin molecules. Hence, purification of pol a removed factor(s) necessary for the activity of the enzyme on B PV DNA. The polymerase activities involved in vivo in BPV DNA replication were analyzed using aphidicolin, a specific inhibitor of DNA pol α, and/or L-canavanine, an inhibitor of protein synthesis. DNA present in infected cells was visualized by autoradiography of Southern blots after probing with nick-translated BPV DNA. Aphidicolin, added at any time after-infection, reversibly inhibited each step of BPV DNA synthesis. Conversely, L-canavanine slowed the replication process, inhibited the synthesis of the viral-coded proteins NP-1 and VP3, and inhibited the production of replicative intermediates (RI) and progeny ssDNA. After removal of L-canavanine, both protein and DNA synthesis resumed. These results demonstrate that 1) pol α is involved in every stage of the replication process including the production of parental replicative form (RF), daughter RF, RI, and progeny genomes, 2) taken in conjunction with the in vitro data, that a pol α holoenzyme complex is required for BPV DNA replication, and 3) viral proteins are required for RI and progeny DNA synthesis, / Ph. D.

Bovine parvovirus DNA is associated with the nuclear matrix of infected bovine fetal lung cells as shown by Southern blot analysis of matrix DNA isolated by two procedures differing in the order of exposure of detergent-treated nuclei to high salt conditions and DNase I. Protein analysis of the two matrix types showed the polypeptide composition to be similar. Both procedures showed enrichment for BPV DNA with progressive DNase I digestion. Over the course of infection the amount of BPV DNA associated with the matrix increased, yet the amount of BPV DNA associated with matrix DNA as opposed to total DNA decreased from 21% at two hours to 7% at eight hours with a subsequent rise to 13% at sixteen hours. Restriction enzyme analysis of the matrix DNA indicated that no specific portion of the BPV genome was responsible for its attachment to the matrix at the selected times. In addition both the nonstructural BPV protein, NP-1, and the capsid proteins VP1, VP2, and VP3 were associated with the matrix at sixteen hours. The association of BPV DNA and proteins with the nuclear matrix implies structural if not functional significance for the matrix in BPV replication. / M.S.

Bovine parvovirus (BPV) is a small icosahedral virus containing single stranded DNA and belongs to the group of nondefective parvoviruses. Research on this virus group has revealed that viral replication occurs in cycling cells only, and evidence exists that the required factor provided by cycling cells is produced during early or mid S phase. Due to this S phase dependence, parvoviruses replicate synchronously only in synchronized cells. This study was initiated in order to determine the kinetics of replication of BPV in highly synchronized cells, the effects of viral replication on host macromolecular synthesis, and the properties of the double stranded DNA produced in infected cells. Bovine fetal spleen (BFS) cells, in which BPV replicates optimally, were synchronized at the G₁/S border by exposure of cells to 2 mM hydroxyurea (HU) for 32 hr. Immediately after release of cells from the HU block by washing, DNA synthesis began. Autoradiographic analysis revealed that within 2 hr, 80 to 85% of the cells were synthesizing DNA. The latent period for progeny BPV production was 8 hr in HU-synchronized cells infected at the beginning of S phase compared to a 16 hr latent period in asynchronous cells. In addition virus titers and the percentage of cells containing viral specific antigens increased more synchronously in HU-synchronized cells than in asynchronous cultures. Synthesis of BPV DNA always preceded the initial increase in virus titers and appeared to govern the rate of virus maturation. BPV DNA synthesis was always initiated during late S phase in cells infected at the beginning of S. Cellular S phase was not affected by BPV replication, but RNA and protein synthesis declined rapidly after the onset of BPV DNA synthesis in infected cells. The double stranded BPV DNA produced during infection was isolated in the supernatant after selective precipitation of cellular DNA with NaCl and sodium dodecyl sulfate (Hirt procedure) followed by hydroxylapatite chromatography. Some of the molecules of DNA isolated in this way existed in a covalently closed circular configuration as demonstrated by rapid reannealing rates following thermal denaturation, In addition, agarose gel electrophoresis revealed bands of DNA in this preparation comparable to ØX 174 closed and open circular replicative forms. No double stranded DNA isolated from cells infected with other animal parvoviruses has been shown to contain covalently closed circular forms. Although BPV is replicated to only a minor extent in stationary BFS cells, BPV was shown to replicate in cells infected 32 hr after the cells were exposed to 2 mM HU but not released. Throughout replication, cells remained in HU and no cellular DNA synthesis was detected. However, BPV DNA and progeny BPV were produced beginning at 16 hr postinfection. This synthesis was not due to viral production in a few cells only, since approximately 75% of the cells were involved in BPV replication as demonstrated by immunofluorescent staining. Therefore, it appears that a factor associated with S phase of the cell cycle and required for optimum BPV replication is produced even in the absence of cellular DNA synthesis per se. / Ph. D.

The termini of the autonomous parvovirus LuIII, which encapsulates plus and minus DNA strands equally, were cloned and sequenced. The left and right termini of LuIII differ in nucleotide sequence and these termini can assume T- and U-shaped intra-strand base-paired structures, respectively. The LuIII termini are virtually identical in nucleotide sequence and secondary structure to those of the rodent parvoviruses MVM and H-1. The presence of non-identical LuIII termini demonstrated that identical ends are not required for the encapsidation of both DNA strands with equal frequency, as suggested for parvoviruses B19 and AAV. An infectious genomic clone of LuIII was constructed and sequenced. The LuIII genome is 5135 bases and it shares over 80% sequence identity with the sequence of the genomes of MVM and H-1. The genome organization of LuIII is virtually identical to that of the rodent parvoviruses of known sequence. The major ORFs, the left and right ORFs, are restricted to the plus strand. Promoter-like sequences are present at map units 4 and 38. The transactivation responsive element (TAR), characterized in H-1, upstream of P38, is also present in LuIII. Regulatory sequences and splice donor-acceptor consensus sequences, characterized in MVM and H-1, are also present in LuIII. This suggests that both LuIII promoters are functional, and that the transcription map for LuIII could be very similar to that of MVM. The LuIII sequence has only a single copy of a repeat present in tandem at the right end of the MVMp genome. Downstream of this sequence, an A-T rich region of 47 nt is present in LuIII. Since this A-T rich region is absent from the genomes of MVM and H-1, we propose that it represents a putative encapsidation signal responsible for the encapsidation pattern observed for LuIII. Northern analysis of BPV RNAs suggests that, like the human parvovirus B19, most, if not all, BPV transcripts initiate at promoter sequences localized at map unit 4. Amplification of BPV cDNA ends by the polymerase chain reaction resulted in a number of BPV-specific fragments. Four of these fragments were cloned and sequenced. Sequencing revealed two splices, one of which is very likely a major splice for several BPV transcripts. cDNA fragments were assigned to transcripts possibly coding for three BPV non-structural proteins. Amplification of BPV transcripts with primers specific to the mid-ORF suggests that the amino terminus of the capsid protein VP1 is not coded for by the mid-ORF as suggested by earlier studies, but instead results from one or both of the two small ORFs present upstream of the right ORF, in the same reading frame. / Ph. D.

The cis and trans signals important in BPV replication were identified using a transient replication assay, the mobility shift assay, and a comparison between the BPV and LPV genomes. Replication of deleted BPV genomic clones, which contain the natural left (3’ OH end of the viral minus strand) and right (5’ PO, end of the viral minus strand) BPV termini, defined the minimum size of the BPV origin of replication (ori) to be the terminal 171 nucleotides of each terminus. Clones containing duplicate termini or altered left ends were also shown to replicate. The BPV ori was determined to have two domains identified by a computer analysis of homologus regions between these termini. Three proteins were identified that bind to the left terminal 171 nucleotides in the hairpin conformation. Inhibition of the formation of the DNA-protein complexes with competitor DNA localized two potential binding sites that correspond to the domains mentioned above. Two of the DNA-protein complexes were formed by BPV-coded proteins as determined by inhibition of the complex by anti-BPV antibodies. The third complex resulted from binding of a host cell S-phase protein that is a likely candidate for the S-phase factor required for autonomous parvovirus replication. The BPV ori thus appears to function by binding both cellular and viral proteins for the initiation of DNA synthesis from the hairpinned termini. The comparison of the BPV and LPV genome sequence suggest that the genomic organization of LPV may be more like BPV than that of the rodent parvovirus minute virus of mice; and therefore, LPV may contain similar cis signals. / Ph. D.